Data against Modern Slavery

20 Jan 2026 14:00h - 14:45h

Session at a glance

Summary

This discussion at the World Economic Forum focused on a new Global Data Partnership Against Forced Labor that leverages agentic AI to combat modern slavery in supply chains. The panel, moderated by Dan Viederman, included executives from HPE and Amazon, the Director General of the International Organization on Migration, and the founder of the Global Migrant Workers Network.

John Schultz from HPE explained that despite numerous well-intentioned efforts, the fight against forced labor is being lost due to information gaps that allow bad actors to hide in darkness. The solution involves using agentic AI to share insights across different data sources without actually transferring sensitive data, enabling organizations to identify risks while protecting confidentiality. HPE has partnered with the Thai government to pilot this system, moving from proof-of-concept to real-world implementation.

Amazon’s Cara Hurst described how AI tools are already helping the company predict high-risk suppliers with 90% accuracy by analyzing audit data, media reports, and geopolitical information. This allows human decision-makers to focus resources on the highest-risk areas rather than spending time on manual data analysis. She emphasized the importance of cross-sector data sharing to improve insights across different industries and supply chains.

Mahendra Pandey, a former migrant worker, stressed that worker engagement is crucial for understanding the real problems and building trust in any technological solution. He highlighted that migrant workers possess valuable data about exploitation through their lived experiences, but they must be meaningfully included in designing systems rather than treated as mere beneficiaries.

Amy Pope from IOM noted that this approach represents a game-changing shift after 25 years of working on trafficking issues, as it can finally penetrate the complex layers that have historically shielded traffickers. The initiative aims to move beyond compliance exercises toward measurable reduction in forced labor incidents through improved detection, prevention, and accountability across sectors.

Keypoints

Major Discussion Points:

– Global Data Partnership Against Forced Labor Initiative: The panel discussed a collaborative effort launched at WEF that uses agentic AI to share data across sectors (companies, governments, NGOs, worker organizations) without compromising confidentiality. The goal is to connect disparate data sources to better identify and combat forced labor in supply chains.

– Technology as a Solution Enabler: Speakers emphasized how AI and agentic AI can transform the fight against forced labor by analyzing patterns across multiple data sources, predicting high-risk suppliers, and enabling faster, more targeted interventions while protecting sensitive information and individual privacy.

– Critical Importance of Worker Engagement: Mahendra Pandey stressed that without direct engagement with migrant workers and survivors, any technological solution lacks “ground truth.” Workers must be involved in the design process to ensure systems are trustworthy, effective, and address real problems rather than perceived ones.

– Moving from Compliance to Impact: The discussion emphasized shifting from traditional compliance-based approaches (audits, reporting) to proactive, data-driven solutions that can prevent forced labor rather than just identify it after the fact. The focus is on creating accountability and measurable outcomes.

– Cross-Sector Collaboration and Transparency: Panelists highlighted the need for unprecedented data sharing between private companies, governments, international organizations, and worker networks, with Thailand serving as the pilot country to prove the concept works in practice.

Overall Purpose:

The discussion aimed to introduce and promote the Global Data Partnership Against Forced Labor, a technology-enabled initiative that leverages AI to combat forced labor through improved data sharing and cross-sector collaboration. The panel sought to explain how this approach could overcome traditional limitations in detecting and preventing forced labor, while building support for broader participation in the partnership.

Overall Tone:

The tone was consistently optimistic and solution-oriented throughout the conversation. Speakers expressed genuine hope that this initiative could be “game-changing” and lead to measurable progress against forced labor. While acknowledging the severity and persistence of the problem, the discussion maintained an energetic, collaborative spirit focused on practical solutions rather than dwelling on challenges. The tone became particularly passionate when Mahendra Pandey shared personal experiences and called for genuine worker engagement, and when John Schultz emphasized the moral imperative to “be ashamed of not looking” rather than being ashamed of finding problems.

Speakers

– Dan Viederman: Schwab social entrepreneur and impact investor focused on creating equitable, inclusive and sustainable supply chains; invests in companies that address forced labor using technology

– John Schultz: COO of HPE (Hewlett Packard Enterprise); has been leading efforts in the Global Data Partnership Against Forced Labor initiative

– Mahendra Pandey: Founder of the Global Migrant Workers Network; former migrant worker who went to Saudi Arabia from Nepal at age 19; leads a network of 23,000 members from 27 countries, primarily low-income migrant workers including domestic workers, with majority being women migrant workers from Africa region

– Amy Pope: Director General of the International Organization on Migration (IOM); previously worked at the Department of Justice on trafficking prosecutions; has been working on forced labor issues for about 25 years

– Kara Hurst: Chief Sustainability Officer for Amazon

Additional speakers:

None identified beyond the provided speakers names list.

Full session report

Global Data Partnership Against Forced Labour: A Comprehensive Discussion Summary

Introduction

At the World Economic Forum, a distinguished panel convened to discuss the launch of the Global Data Partnership Against Forced Labour, an innovative initiative leveraging agentic artificial intelligence to combat modern slavery in global supply chains. The discussion was moderated by Dan Viederman, a Schwab social entrepreneur and impact investor focused on creating equitable and inclusive and sustainable supply chains who invests in companies, including those that address forced labor using technology.

The panel featured John Schultz, Chief Operating Officer of Hewlett Packard Enterprise, who has been spearheading the technical development of the partnership; Cara Hurst, Chief Sustainability Officer for Amazon, representing corporate implementation of AI-driven risk detection; Amy Pope, Director General of the International Organisation on Migration, providing institutional perspective from 25 years of anti-trafficking work including prosecutorial experience at the Department of Justice; and Mahendra Pandey, founder of the Global Migrant Workers Network and former migrant worker, offering authentic worker voice and advocacy experience.

For those interested in more technical details, a white paper was launched by the World Economic Forum that morning and can be found by searching for “Global Data Partnership Against Forced Labor.”

The Global Data Partnership Initiative: Technical Foundation and Vision

John Schultz positioned the Global Data Partnership Against Forced Labour as a response to the fundamental challenge that despite numerous well-intentioned efforts, the fight against forced labour is being lost due to critical information gaps. He explained that these gaps create darkness where bad actors can hide and exploit vulnerable workers within complex supply chains.

The technical innovation at the heart of the partnership centres on agentic AI, which Schultz described as the “secret sauce” that enables organisations to extract insights from shared data without actually transferring sensitive information. This approach addresses the historical barrier to cross-sectoral collaboration: the need to protect confidentiality whilst enabling effective information sharing.

Schultz provided a compelling example of how this technology would work in practice: “If a survivor shows up on the front lines and says, you know, I just freed myself from bondage at this location, and the people who were involved in my recruitment, and essentially my enslavement, was this person and this person, and on my end of my supply chain team, I can be doing an agentic crawl that sees the name and sees the location, and indicates that it’s actually someone we use in our supply chain. I don’t need to know who the survivor was, I don’t even need to know their circumstances, I don’t have to jeopardise their personal confidentiality, but I can immediately take action on my side.”

The initiative has progressed from proof-of-concept to minimum viable product stage, with Thailand serving as the pilot country for full production implementation. The partnership includes multiple collaborators, as Schultz noted: “we have a lot of partners, AWS, Cisco, IOM, and the like.”

Schultz expressed confidence that the corporate economy could achieve material reduction or complete eradication of forced labour through eliminating the information gaps that currently enable exploitation. However, he provided an important clarification about scope: “What I mean by the corporate economy is where you have corporate supply chains and where traditional organizations and governments are engaged. When you’re into the black market and things like that, right, that’s going to be a different level of things.”

Corporate Implementation: Amazon’s AI-Driven Approach

Cara Hurst detailed Amazon’s existing implementation of AI tools for forced labour detection, demonstrating the practical application of predictive risk analysis in corporate supply chains. She explained that Amazon has built “an AI tool inside of Amazon, which is predictive risk analysis” that can identify high-risk suppliers with remarkable effectiveness. As Hurst specifically stated: “nine out of 10 times, it’s found the highest risk suppliers” by analysing audit data, media reports, and geopolitical information.

This technological capability allows human decision-makers to focus their resources on the highest-risk areas rather than spending time on manual data analysis. Hurst emphasised that technology should augment human decision-making rather than replace it, maintaining humans in the loop for all critical decisions whilst leveraging AI’s analytical capabilities.

Hurst drew an important comparison to carbon tracking work, noting: “Those of us who work on human rights primarily are always a little envious of the progress that people in focus on the carbon have made.” She explained how actual data versus modeled data enables faster business decisions and more effective interventions.

The Amazon experience illustrates the potential for cross-industry data sharing to accelerate learning and improve risk identification across different sectors. Hurst noted that sharing aggregated data across industries could significantly enhance predictive capabilities and enable more targeted interventions.

Furthermore, Hurst highlighted how real data insights could drive policy advocacy initiatives, moving beyond individual company responses to collective industry action for systemic change. She stressed the importance of a human-centred approach that ensures technology serves people’s lives rather than becoming an abstract data collection exercise.

International Organisation Perspective: Historical Context and Transformation

Amy Pope provided crucial historical context, positioning the current initiative within her 25 years of experience working on trafficking and forced labour issues, including her background in trafficking prosecutions at the Department of Justice. She described the evolution from initial denial of the problem’s existence to current recognition and the potential for technology-enabled action.

Pope characterised the Global Data Partnership as a “game-changer” that could fundamentally remove the shield that traffickers have been using to hide their activities. She explained that the approach represents a paradigm shift because it can penetrate the complex layers that have historically protected traffickers and enabled exploitation of vulnerable populations.

From the International Organisation on Migration’s perspective, the partnership offers unprecedented capability to understand migration flows, identify trafficking networks, and enable more effective interventions. Pope emphasised that prevention through border management and worker education is equally important as detection and remediation efforts.

She provided specific examples of preventive work: “we have started educating migrant workers in the east of Africa who are going to work in the Gulf about what are their rights, what are the expectations. Just that little bit of knowledge and knowing where to go, knowing who to call when things go the wrong way, has made a significant difference.”

This preventive approach, combined with improved detection capabilities, could significantly reduce the number of people entering forced labour situations.

Worker Voice and Authentic Engagement

Mahendra Pandey brought a fundamentally different perspective to the discussion, grounded in lived experience as a former migrant worker who travelled to the Gulf region in his early 20s. He provided concrete details about the exploitation he experienced, including paying “55,000 Nepali rupees for the recruitment agency” and working “in the heat temperature, like, you know, 45 to 50 degree temperature in the Middle East and Gulf” with accommodation conditions where “migrant workers are staying in the one room, eight people, nine people, sometimes without air conditions.”

Pandey’s advocacy work now spans a global network representing “27 countries, and most of whom are low-income migrant workers, including domestic workers, and the majority of them are women migrant workers from Africa region.” He noted that “there are more than 37 million migrant workers in the Gulf and Middle East,” providing crucial context for the scale of the challenge.

His interventions consistently challenged the panel to centre worker voices and experiences in solution design. Pandey emphasised that “data is not just a number” but represents people’s lives, journeys, experiences, and stories. He argued that without worker engagement, it is impossible to understand the real problems, and without understanding the problems, effective solutions cannot be developed.

His most significant contribution was highlighting the trust dimension: “Without worker engagement, if there is any AI, any tools is built up, and worker is not going to trust. And if, in the design process, if worker are, you know, able to be part of the design process, then they also become responsible to trust, and also implement, and at the same time, share the truth.”

Pandey challenged assumptions about migrant workers’ technological capabilities, noting that they possess technological sophistication and learn faster than expected. He stressed that migrant workers possess valuable data about exploitation through their lived experiences, but they must be meaningfully included in designing systems rather than treated as mere beneficiaries.

He also provided a broader critique of resource allocation priorities, questioning why society invests significantly in entertainment infrastructure whilst migrant justice receives comparatively limited resources. This perspective highlighted the need for systemic change that addresses the complete spectrum of worker exploitation and trauma.

Areas of Strong Consensus and Key Tensions

The discussion revealed remarkable alignment across multiple dimensions despite the diverse backgrounds of the speakers. All participants agreed that data sharing and cross-sectoral collaboration are essential for combating forced labour effectively, recognising that isolated efforts are insufficient and that shared insights lead to more effective outcomes.

There was strong consensus that technology should augment human decision-making rather than replace it, with both corporate and advocacy perspectives emphasising the critical importance of maintaining human oversight and engagement in AI-powered systems.

The speakers unanimously agreed on the fundamental importance of worker engagement and education, recognising that workers must be educated about their rights and actively involved in solution design, as systems cannot capture ground reality or build necessary trust without worker participation.

However, notable differences in emphasis emerged between speakers. The primary tension was between technology-focused and human-centred approaches to solution design. John Schultz emphasised the technological solution of agentic AI as enabling data sharing without compromising confidentiality, whilst Mahendra Pandey stressed that technology without worker engagement in the design process cannot capture ground truth or ensure system effectiveness.

There were also differences in the scope of achievable impact. Schultz expressed confidence that forced labour could be materially reduced or eradicated from corporate supply chains, focusing on corporate contexts. In contrast, Pandey argued for broader systemic change addressing the complete spectrum of worker exploitation and trauma as necessary for true prosperity.

Implementation Challenges and Unresolved Issues

Despite the strong consensus on the initiative’s potential, several critical implementation challenges remain unresolved. The specific technical details for maintaining data privacy whilst enabling effective sharing require further development, particularly as the system scales beyond the Thailand pilot.

Mechanisms for ensuring worker safety and preventing retaliation when they report violations represent a crucial unresolved challenge. The success of the entire initiative depends on workers feeling safe to share authentic information about their experiences.

Scalability presents significant challenges for expanding beyond Thailand to multiple countries simultaneously. Different regulatory frameworks, cultural contexts, and institutional capabilities will require careful navigation.

Resource allocation and funding models for sustained global implementation remain unclear. The initiative will require substantial ongoing investment to maintain effectiveness and expand reach.

Integration with existing regulatory frameworks and compliance systems across different jurisdictions presents complex technical and legal challenges that must be addressed for successful implementation.

Measurement and Accountability Framework

A critical theme throughout the discussion was the need for measurable outcomes rather than activity-based metrics. Dan Viederman consistently pushed for accountability mechanisms that could demonstrate actual reduction in forced labour incidence rather than simply measuring participation or data sharing.

The speakers agreed that the ultimate success metric should be measurable reduction in forced labour incidents, but the specific methodologies for tracking such outcomes remain to be developed. This represents a significant challenge given the hidden nature of forced labour and the difficulty of establishing baseline measurements.

John Schultz proposed a fundamental shift in accountability frameworks, arguing that companies should reframe their approach: “Don’t be ashamed of what you’re finding. Be ashamed of the fact that you’re not looking. Be ashamed that you’re not part of the data partnership.” This reframing could create powerful incentives for proactive participation in transparency initiatives.

Future Vision and Scaling Potential

The discussion concluded with an optimistic vision for the initiative’s potential impact. The speakers expressed genuine hope that this approach could achieve material reduction or eradication of forced labour from corporate supply chains through systematic elimination of information gaps.

The Thailand pilot will serve as proof-of-concept for scaling to additional countries and partners. Success in demonstrating measurable impact in Thailand will be crucial for building broader support and participation.

The initiative aims to move beyond traditional compliance-based approaches toward proactive, data-driven solutions that can prevent forced labour rather than simply identifying it after exploitation has occurred. This preventive focus, combined with improved detection capabilities, could fundamentally transform the landscape of worker protection.

The speakers envisioned expanding data sharing partnerships across sectors to improve predictive capabilities and enable more targeted interventions. Policy advocacy initiatives based on aggregated data insights could drive systemic changes beyond individual company responses.

Dan Viederman concluded with a call to action: “If you want to learn more, go to the website, send us a note, join, be part, be willing to share your insight, share your data, share your tools. We’re in open process right now.”

Conclusion

The Global Data Partnership Against Forced Labour represents a potentially transformative approach to combating modern slavery through technology-enabled collaboration. The discussion revealed strong consensus among diverse stakeholders on the initiative’s potential whilst highlighting critical implementation challenges that must be addressed.

The success of the partnership will depend on maintaining the collaborative spirit demonstrated in the discussion whilst addressing technical challenges of cross-sectoral data sharing, ensuring worker safety and authentic engagement, and developing robust measurement frameworks for accountability.

The initiative’s emphasis on transparency, worker engagement, and measurable outcomes positions it as a significant departure from traditional compliance-based approaches. If successfully implemented and scaled, it could indeed represent the paradigm shift that Amy Pope described as a “game-changer” in the 25-year fight against trafficking and forced labour.

The path forward requires continued commitment to the principles articulated in the discussion: technology that augments rather than replaces human decision-making, authentic worker engagement in solution design, transparency that eliminates information gaps, and accountability for measurable reduction in forced labour incidents. The Thailand pilot will provide the crucial test of whether these principles can be translated into effective practice at scale.

Session transcript

Dan Viederman

Hello, everyone, and welcome to this afternoon’s session, both people who are here with us in person and online. My name is Dan Viederman. I am a Schwab social entrepreneur and an impact investor focused on creating equitable and inclusive and sustainable supply chains.

We invest in companies, including those that address forced labor using technology. So very much consistent with the theme for today. On the table today, among other things, is how can agentic AI, all the rage these days, help us in the fight against forced labor?

We’re not going to talk too much about the problem of forced labor. If you’re here, if you’re there, there are lots of resources that can help you understand it. What we want to talk about is a solution that is emerging.

It’s emerged over the last couple years, started in a conversation here at WEF a couple years ago, and then launched formally last year at WEF. And the question really is, how can we bring together disparate data sources? How can we bring together different pieces of information and different stakeholders around the same digital table, and thereby increase our ability to understand the reach of forced labor, both in supply chains and elsewhere?

How can we mobilize our own knowledge, and fundamentally, how can we increase the amount of action that we take? How can we create accountability for ourselves and those of our, for our partners? For me, ultimately, this is about creating a system, early stages that we’re at, that can lead to progress against the persistent problem of forced labor.

We have a wonderful panel with us today. John Schultz, who’s COO of HPE, and has been very important in this effort. Mahendra Pendi, who’s the founder of the Global Migrant Workers Network.

Amy Pope, the Director General of the International Organization on Migration. And Cara Hurst, Chief Sustainability Officer for Amazon. John, let’s start with you, as someone who’s been with us for a couple years leading this and leading the conversations both internally at HPE, with the WEF, and with other businesses.

What’s on the table? Why is this initiative going to give us the potential to make progress in a way that others haven’t?

John Schultz

Well, thanks, everyone, for participating. You know, when we looked at this problem a couple years ago, what we concluded was there are a lot of people engaged in the effort to eradicate forced labor, and yet we’re losing the battle. And so fundamentally, the question was, how can we have all these people, well-intended, doing an incredible amount of work on the ground, and we’re still losing ground?

And what we concluded was, the solve is we’ve got to find a way to make everyone who’s engaged more effective. And as a technology company, what we know is that data leads to insights, and insights lead to more effective outcomes and actions. And so the question was, okay, how do we generate more insights for the people who are engaged in this fight so they can drive better outcomes?

And that meant sharing data. We have seen any number of instances in which data sharing has allowed us to have a better outcome. Maybe one of the most powerful, if not the most powerful, in the last few years was the ability of people in the vaccine world to get together and share intelligence to find vaccines for COVID.

And we supported that effort with our supercomputing, et cetera. So you bring the infrastructure, you bring the data, and you work together, and you accelerate the time to value. And so what we concluded here was that if folks on the ground who have data from survivors and the like can marry their data with governments who are involved in lots of different activities, along with NGOs who are focused on immigration and migrant rights, with companies who focus on supply chain, if we can make all that happen, everyone in the chain can be better.

Agentic AI is the secret sauce, because in the old world we would have had to do all these special data structures and then somebody would be like, well you got to read my data, and I got to read your data, and I’m worried about confidentiality, and I’m worried about my own, you know, my own business propriety, etc.

The beauty of agentic AI is we can pull insights out without actually having to transfer the data. So the example I like to use is, if a survivor shows up on the front lines and says, you know, I just freed myself from bondage at this location, and the people who were involved in my recruitment, and essentially my enslavement, was this person and this person, and on my end of my supply chain team, I can be doing an agentic crawl that sees the name and sees the location, and indicates that it’s actually someone we use in our supply chain.

I don’t need to know who the survivor was, I don’t even need to know their circumstances, I don’t have to jeopardize their personal confidentiality, but I can immediately take action on my side to go look at that site, to go talk to those people, and see if I can’t change the game.

So agentic AI is the secret sauce in this, and that’s what’s kind of given rise to this global data partnership, where we’re trying to get people to share data with an agentic AI overlay, and make everyone in the fight more effective.

Give us another word or two on how far we’ve gotten, what’s been built so far, and what does it look like? Two answers to that as a corporate guy. We’ve gotten so far, we have not gotten nearly far enough, right?

I mean, that’s right, it’s the healthy impatience. So we’ve stood up the architecture, we have a lot of partners, AWS, Cisco, IOM, and the like, so great partners, and we are moving from the proof-of-concepts stage to the MVP stage. But again, thinking about, thinking as a corporate guy, activity isn’t really what matters, outcomes are what matters.

And so what we’ve challenged ourselves to do is find one country where we can bring this solution and show that it makes a difference. And fortunately, we’ve gotten tremendous support from the government of Thailand. That partnership is critical because without government, cracking this problem in a particular location would be incredibly hard.

So for year two, we’ve got to move this into full production. We have to do it in Thailand, and we have to show that it works. And if that is achieved, next year we’ll be back here talking about how many more governments can we enlist in our effort, how many more partners can we get, and therefore how much more we can scale it.

So year one, tremendous progress. I think we’re on schedule. Year two, we now need to prove it works in Thailand, and then we’ve got to go faster.

Dan Viederman

For those interested in more of the technical details, there’s a white paper online that was launched this morning by WEF. I think if you Google Global Data Partnership Against Forced Labor, you’ll be able to find it. And it goes into some more detail about the technical architecture.

But the concept, of course, is improving and increasing cross-sectoral collaboration with irrefutable data. That is also protected. Cara, Amazon, it’s a large company.

You’re in most sectors around the world. Obviously, the company has developed systems and processes internally that embrace and address forced labor for your own operations. You’ve also been incredibly important in supporting some cross-sectoral collaborations.

As you think about the role of data in particular and this initiative, how will it help you prioritize and take action?

Kara Hurst

Yeah. Well, I think that was a great introduction to a lot of what we think is a huge opportunity. We’re really excited about what AI can do to accelerate our ability to take what were very disparate signals before and connect those insights.

I’ve been working on this for a long time. Our teams have been working on this. We were just talking about how far back some of this work goes.

And I think this is just, it’s a really critical catalytic moment right now where the technologies that we’re building, I do have real hope that they’re going to transform how we’re able to look at risk, how we’re gonna be able to make those insights actionable, and most importantly, how we’re going to be able to change people’s lives.

Because all of that, we can talk about data, we can talk about information, but what is it going to mean for people? So I wanna make that, keep coming back to that connection as well, because we can talk about these things in the abstract, and we always talk about this internally, talking about risk and talking about data and collection and open data and all of that is good, but what is it going to mean for people on the ground?

So I do wanna come back to that at some point, but some of the things that we’ve been thinking about at Amazon when we think about data at scale and experimenting with, and the value of these public-private partnerships and collaboration, we have hundreds of thousands of suppliers around the world globally, and we obviously have a supplier code of conduct, and we expect all of our suppliers to adhere to that, we have a global human rights set of principles, all of this is public, we are open about sharing policies, improving them, learning from others, all of that, but the monitoring of that, the capability building of our suppliers, their capacity to respond and all of that is challenging, we know that.

So what we want to do is to try and learn as fast as we can to go in to understand where the highest risks are to remediate and to continue to hold that bar high. So what AI has been allowing us to do, we’ve built a couple of, just to get very specific about it, because I think to get tangible too helps, we’ve built an AI tool inside of Amazon, which is predictive risk analysis, and as we’ve started to look at that, what it does allow us to do is to really focus our resources on where the highest risk areas are.

What it first allows us to do is to kind of go in and scan the, analyze the historical audit information that is available and kind of take a look across of that, what we have available. Then it also allows us to, using computer-generated simulated information to kind of pull from other types of audit information that are out there. And then we have an ability to kind of look across media reporting, analytics, geopolitical data, which would give us information about what would the particular, what would we think about the risks for the supplier base where we’re looking.

And all of that put together, I think also to your point, protects confidentiality and it protects, you know, the specifics of any particular individual or even like, you know, factory relationships to some extent.

The other thing it does is it really just helps us to not, it’s not going to ultimately make decisions. And I want to make that distinction about where we use AI and where we don’t. All of this is information, but then a human, we’ll still call human in the loop when we’re using AI tools.

So at the end of the day, AI is just going to give us really good predictive information and analytics, but you still have a human in the loop at the end of the day making all of our decisions. And I think that’s a very important distinction as we talk about these tools. But the interesting thing as we’ve gone through and used this PRISM tool to make this predictive risk analysis, is when we’ve gone back and looked, nine out of 10 times, it’s found the highest risk suppliers.

So when we’ve gone back and we’ve looked at this, is it working? It is. And what it allows us to do is instead of utilizing all of our team’s time to go in and do that heavy lifting of going through these audit reports and looking and spending all the time doing this risk analysis, we have better risk analysis, much, much quicker, with much more thorough information, and our people can spend their time.

doing the decision-making and hopefully more of the capacity building, the capability building, the engagement with the suppliers or getting the worst of the suppliers out and working with the best of the suppliers.

So I think what it’s going to help to do is start to improve those relationships with supply chain and again connect those signals and focus us on the important work.

Dan Viederman

I mean, right, the process of trying to understand where the risk lies in your supply chain and really any company, much less one that’s as massive as Amazon, is a clunky, old-fashioned process that involves social audits and other things we don’t have to talk too much about.

But this has the promise of short-cutting the things that can be short-cut and putting decision criteria in the hands of individual people. So let’s absolutely come back to a question about kind of what that decision-making process looks like. But in the moment, Mahendra, I’ll turn to you.

One of the other things that’s difficult for companies within supply chains is sort of engagement with the people who do the work in the supply chains, workers, right? You, yourself, a migrant worker at age 19, went to Saudi from Nepal and now founder and facilitator, what do you call yourself? Orchestrator?

Scaffolder of the Global Migrant Workers Network, which has 23,000 members around the world. The opportunity to engage with workers throughout the world due to your network puts those in the private sector and government in a different position. How, in particular, do you see participation of your members and your network in this global data partnership?

Mahendra Pandey

Sure. Thank you so much and great to be here and being a migrant worker in Saudi Arabia in early 20s and now coming to the World Economic Forum and being a Davos, it was, you know, life-changing experience for me, but I wanted to also make sure that, you know, all the migrant workers like me, they deserve a similar kind of journey.

So there are more than 37 million migrant workers in the Gulf and Middle East and who have been working there. Of course, everybody is not here, but I am one of them. So, you know, why sometimes when we talk about data partnership and data and AI, and data is not just a number, and we are talking about, you know, people’s life and their journey, their experience, and their story as well.

So if you are wondering about, you know, what kind of data, we are talking about migrant workers, you know, their passport is taken away, and their salary is, you know, delayed, and they are scared about, you know, being deported, and they don’t have proper accommodations, and they have to pay high recruitment fee.

Like me, when I went to Saudi Arabia, I had to pay 55,000 Nepali rupees for the recruitment agency. And if I didn’t have to pay that, you know, that money I could have sent to my mother. And you know, if the recruitment fee you have to pay for, you know, your money lenders and your recruitment agency, then, you know, most of you are six months, one year, and you have to work for free to pay your loan back.

And we wanted to make sure that, you know, whatever discussion we are talking about data partnership or AI and technology, and we wanted to make sure that that system that helps to reduce or like, you know, solve those problems.

And why, you know, this partnership is so important for migrant worker like me or migrant worker-led organizations like ours, because without worker engagement, you cannot understand what is the problem.

And without understanding the problem, you cannot solve the problems. So you know, sometime when you develop the AI or technology, without worker engagement, and you cannot have the ground truth, and you cannot have the ground reality. And also, whether it is going to work or not, and without worker engagement, you are not going to know.

That’s why our participations and our engagement with this initiative is very crucial. And also, we wanted to, you know, build up the trust, and there is a stereotype. among government and private sectors, and if we invite migrant workers to the stairs like this, they always hear sad story, bad story, they cry.

But if migrant workers are, like, you know, being deported and their salary is not, you know, being paid, their passport is controlled by employer, and they are not given proper break, and they have to work, like, you know, in the heat temperature, like, you know, 45 to 50 degree temperature in the Middle East and Gulf.

And what else they can see here? Of course, those are the reality. But at the same time, if you are celebrating, and if you are designing the, like, you know, AI technology, and you have to have also courage to listen migrant workers, you have to have also courage to listen the, like, you know, survivors, and also their story, their experience, their trauma, and so that those things, you know, you can build up while, you know, designing the technology and the system as well.

So, you know, sometimes there is also a stereotype about, you know, migrant workers are illiterate, and they don’t know technology. But if you go to the, like, you know, some of the labor camp or, like, you know, migrant worker community, they live in the cyber world, and they know, like, you know, all the technology. They learn faster than maybe, you know, each of us in this room, because they are desperate, and their courage, and their, like, you know, happiness, or the joy they bring is, you know, extraordinary.

And that’s why, you know, Dan mentioned about the Global Migrant Worker Network is the survivor-led and purely worker-led from the Global South, from the 27 countries, and most of whom are low-income migrant workers, including domestic workers, and the majority of them are women migrant workers from Africa region.

This is not just a, like, you know, platform for beneficiary and talk and share about, like, you know, sad and bad stories. It’s the platform about, you know, sharing about some ideas, and also sharing about the solution, engaging the initiative like this, and making sure that, you know, we have a voice. And without worker engagement, if there is any AI, any tools is built up, and worker is not going to trust.

And if, in the design process, if worker are, you know, able to be part of the design process, then they also become responsible to trust, and also implement, and at the same time, share the truth. And those kind of things, I wanted to share here. And one thing, you know, I want to humbly request everyone, you know, trust survivor, trust migrant workers, and engage with them.

And if we engage with the migrant worker and survivor, any kind of, you know, system we build up, that system is going to be

Dan Viederman

work perfectly. Thank you so much. Thanks, Mahindra.

Yeah, there’s obviously a practical reason to engage with workers in the sense, as you suggest. Otherwise, you don’t really know what the problem is. There’s also a moral reason to engage with people who are most affected by the decisions some of the rest of us are taking.

I think a lot of people, when they think about or hear about tech, AI, as it relates to individuals, maybe do have that impression of sort of lack of sophistication, a lot of risk, as you’re suggesting.

To the extent that migrant workers or workers in general are involved in the design of the technology, then the risk can be mitigated. But of course, there’s also continually concerns about confidentiality. As John mentioned, the system is designed to respect the sovereignty, the confidentiality of a data holder, be it an individual or an organization going forward.

So, an important design consideration as well. Last but not least, Amy, come to you. You sit within one organization that operates across sectors.

You yourself have been involved back in the Department of Justice years ago in trafficking prosecutions. How do you see the opportunity, particularly, to link data across sectors and increase impact and increase outcomes and increase accountability?

Amy Pope

It’s a game-changer. So, put it this way, I’ve been working on the issue for about 25 years. 25 years ago, people didn’t actually believe that there was still modern day slavery.

If you told your average consumer, if you spoke to your average CEO, people didn’t recognize it. So we start there, we started with the education, and then we got to a place where everyone agreed it’s a bad thing, it’s not good for markets, it’s not good for consumers, it’s not good for companies themselves, it’s not good for workers, but the ability to detect it was increasingly impossible because of the complexity of global supply chains, because of the ways in which companies use subcontractors, and because of the ways that unscrupulous recruiters and traffickers were manipulating systems to basically hide within traditional ways of working.

And so I really want to say thank you to the forum and to those of you who are sitting on the stage with me, and importantly to the migrant workers themselves, because the way that we are now taking on the issue is going to fundamentally remove the shield that traffickers have been using to hide their unscrupulous activities on the backs of some of the most vulnerable and desperate people in the world.

And that’s really the magic of what we’re seeing here. Historically, even the last 10 years, there’s been a lot of interest in how do we use data. I’ve worked with various companies who’ve looked at their own supply chains, they’ve asked for help in building out ethical recruitment processes, they’ve looked for patterns that have identified where there are bad actors in the system, but it was nearly impossible to get all the way down to the recruitment of the worker, him or herself, because there were so many levels in between, and it was not reasonable to ask someone like John to tell me what’s happening within a factory in X, Y, or Z country.

Because John would say, look, I’m looking at my supply chain, it looks good, everybody’s checked off, but beneath the surface, there was tremendous exploitation happening. So what we’re doing now, being able to safeguard data but to share it, having the input of the migrant workers themselves, who are seeing and experiencing it in real time, having a partner like Thailand, who’s all in in coming up with a solution, and having some key partners who are willing to step forward and say, look, this matters to me, we’re gonna put some skin in the game and we’re gonna figure out a solution.

This could really change the dynamic in ways we have not seen since we’ve recognized the

Dan Viederman

phenomenon. Wonderful. Very optimistic, I appreciate that.

Let’s turn towards sort of what we expect to change at a sort of a level of specificity, or at least hope, I guess. So Cara, let me start with you. I mean, as you think about this initiative, you know, how far can we get, I guess?

I mean, what would credible progress look like in the future? What would Amazon want to see from this? How would it help you make decisions differently?

And what decisions

Kara Hurst

would those be? I think one of the things that will be really transformative when you talk about that vision for the future is the sharing of data. The more that people are willing to share data and do that in aggregated ways, and we can facilitate that through this open data partnership, and we can share what we know, we can gather new insights, all of the models that we’re building, all of the, you know, ways in which we can aggregate information, the more that we can share, the more we know, the more we know across industries.

And even with, you mentioned that we’re in multiple industries at Amazon. We’re certain in multiple different industry segments. Agricultural commodity supply chains look very, very different than electronic supply chains, which again look different than the healthcare industry, which looks different than aviation or maritime.

And we work in all of those. And so what I know about a supply chain in one part of my business is going to be very different than another. And while the lessons are transferable, the information about those supply chains and how workers are treated, and the migration patterns, and then you overlay geography onto that.

And it’s a very complex pattern around the world. of what’s happening in the intersection between workers and what traffickers know and how sophisticated they’re becoming as well with technology. We’re not the only people implementing technology, unfortunately, and what’s happening in those industries.

So I think the more that we can share information and be open about it and say, this is an area where it is pre-competitive, it is an area where collaboration allows us to go much faster, and if we can find those ways, and this is certainly an area of your expertise, Dan, but where we can find ways to aggregate and share data in an open setting, we now have the technological tools in things like AI and agentic AI that will allow us to even go further, and then quantum, which is gonna even allow us to drive new insights in the future, that those kinds of tools that we are now seeing and building and testing and trialing and will continue to build will be informed the more data that we have.

And if I take the work that we’ve done in carbon, which we are a little bit further ahead in, to be honest, and I think about that as an analogy, the information that we have as actual inputs versus what we’re able to model has been really an interesting path that we’ve been walking on for a long time.

I can have certain inputs in my business where I know the actual numbers, and when I have an actual piece of information about a carbon emission versus I have to model an economic input-output analysis, I can do things very differently in the business.

I can drive a decarbonization decision much faster with a business leader than I can when I just have to model something. So it’s great that we can model certain pieces of information, but when you have actual information, I think you start to also say, we know this information, we’ve put it together, we have these actionable insights now.

And then you can go and say, what are the policy levers we need to then come together and drive and we can advocate for in a collective way as an industry? What are the policy changes we need to see that will be better for workers? We can go and talk to workers and say, what would benefit you?

Because of course, community always knows best. So I think those kinds of things, once we have real information, we now have the ability to drive the insights much faster and get those changes going. So I’m excited about that, but we need that collective ability and trust to find those data sources.

Dan Viederman

Those of us who work on human rights primarily are always a little envious of the progress that people in focus on the carbon have made, which isn’t to suggest that we’re unhappy with progress against carbon, but rather the issues themselves are different.

One is sort of less measurable, less easily measurable, less scientific, it involves human activities. I guess one of the things we’re talking about, and maybe there’s an analogy to the carbon world as well, is that we really have to talk both about remediation, that is identifying where issues happen, and prevention.

You mentioned, Cara, the opportunity to engage policymakers as a sort of a preventative approach. But maybe I come to you with a version of the question that I asked Cara, which is sort of, what progress do you think we can credibly make? And in particular, as someone who’s operating at the sort of regulatory level, but at the intersection of government and business, how do you see this leading in the direction of more preventative activities, not merely remedial activities?

Amy Pope

So it’s key that the insights that are pulled out of this are shared with government actors. Ultimately, for example, we at IOM have partnerships with governments to help them better manage their borders and understand who’s crossing their borders. And we’ve been able to use AI to help various governments who may not normally have the kind of insight to understand what are the flows of people and what are they, basically, what are they up to?

Not down to the level of the individual, but to understand, well, this trafficking network has been operating across our border. So the more that we can share the insights that we get from something like this with our government partners, with those who are implementing on the front lines, the more they can take steps to manage their borders.

Secondarily, a key piece of this is educating workers before they go. Because workers are often sold a pipe dream. They’re often told you’re gonna have a fabulous job, you’re gonna be able to send money home, yes you just have to pay this fee.

They don’t really appreciate what they’re walking into. So the more we can use it as a way to educate communities before they go and work abroad, the better outcomes. And then finally, once and the more we’re able to empower workers to report out and then ensure that there’s capacity within governments to respond to it, the better outcomes we’ll get.

So the private sector is key to make sure that they’re setting the standard and they’re setting the expectation, right? Because if everybody is happy to look away, then nothing is going to change. So having the private sector set the demand signal, having governments come behind with these are regulatory expectations, and then helping to fill in the gaps so that governments, community leaders, migrant workers themselves have the capacity before especially people take that job, the better outcomes we’ll get.

Dan Viederman

So you could see IOM being part of this pulling in some of the data and sharing it with both your governmental partners, any businesses that ask you questions, in your engagement directly with workers in-country as you have.

Amy Pope

Absolutely. And we see really good impact when we do. For example, we have started educating migrant workers in the east of Africa who are going to work in the Gulf about what are their rights, what are the expectations.

Just that little bit of knowledge and knowing where to go, knowing who to call when things go the wrong way, has made a significant difference.

Dan Viederman

Mahindra, former migrant workers, someone in touch with migrant workers all the time, what are your hopes for this? What practically do you think? this can lead to, such that it enhances the work life, the quality of life for the people who are part of your network and who will continue to join your network as you grow?

Mahendra Pandey

Sure. No, I agree with you, Amy, that worker needs to have educations and also they deserve to have educations and access to information, and that helps a lot, and not being in expectations. But unfortunately, even though there is a number of education and awareness being done by the, like IOM and ILO and the NGO and some private sectors in the government, those are not enough, and we need to put more resources.

And just a simple example, if we are able to invest resources to build up the studio on Disneyland and other thing, why cannot we put the resources to have the migrant justice, why cannot we put the resources to have justice for the people, and ultimately everybody is going to be benefited for that.

And some countries, they are developing and they talk about number of GDP, number of economy, and also infrastructure, like building towers, bridges, and roads. First, all those are being built up by the migrant workers, but at the same time, when people they see, they only see the beauty of like towers, buildings, studios, Disneyland, those kind of things. And how about like exploitations?

How about salary is not being paid? How about passport is taken away? How about migrant worker are being deported?

How about they are being surveillance? And if we able to make that darkness in the beauty, there will become complete kind of a beauty that migrant workers, unless there is exploitation, unless there is a sadness, unless they have to leave their home country, nobody wants to like, you know, leave.

Unless they have to, you know, for me, I had to leave because of the poverty and I wanted to, you know, help my family, my dad also went for the same purposes, millions of migrant workers also they are doing same thing.

And one important thing, you know, I wanted to make sure that all of us recognize and also understand that, you know, without addressing those exploitation abuses, and also that trauma that worker have, there won’t be the like, you know, complete prosperity, we talked about, you know, in the Davos prosperity dialogue and those kind of without their engagement without understanding them, it won’t be that.

And for me, you know, this partnership is very unique. And since I joined, you know, I believe and I appreciate the collaborators and WAVE and SP and other in the room. This is not just the tokenism, it’s giving power, giving agents to the workers.

That’s how I believe and that’s what, you know, I wanted to contribute as well. So why worker, you know, they need to be a part of that, you know, when we talk about worker exploitations, these are the like, you know, evidence. And when we talk about worker story, these are the data.

So if you look, talk with the, you know, millions migrant workers in the Gulf and network like, you know, ours. And if you want to hear more story about recruitment fee, more story about exploitations, more story about like, you know, whether that company is able to address some of the issue or not, you can get the data, but you need to make sure that you have a trust.

And how do you build up the trust with the workers, and you need to make sure that those data are private and their privacy is secured. And, you know, by saying their story and when workers say their experience and making sure that they are not deported, they are not losing job and making sure that, you know, they are empowered enough and they build up the trust with the private sectors and governments.

And that is second thing. And third thing is, when you talk about data, and if you only look at the data about company audits, or like UN reports and other things, you cannot get the data from audit findings. And if you look at the audit findings of the company, they are not going to say that equipment fee abuses.

And they are not going to say that workers are deported. And they are not going to say that migrant workers are staying in the one room, eight people, nine people, sometimes without air conditions, and those kind of things. In order to have real data and real story, you need to engage with the workers.

In order to engage with the workers, you need to invite and welcome them. And we want to also welcome you as well. Not only you welcome us, we also welcome you.

Come us, visit us. Come us, Kenya. Come us in Qatar, Saudi Arabia, Nepal.

And be with us, and see our life.

Dan Viederman

Wonderful, and obviously the data partnership is one way of kind of virtually and digitally engaging one another, even if we can’t all go visit your network where you gather on a biannual basis. John, back to you. I think a sort of two-part version of this question, one of which is there’s always a danger in anything where there is regulation and the potential for embarrassment, whether you’re a corporate or an individual.

And certainly forced labor is an area where there is kind of shame attached to it, and also regulatory penalty. And so how do we, first part, how do we make sure this does not turn into another compliance exercise? We all have lots of those.

We don’t need to do more work to make sort of compliance tools, right? They exist. I guess the second part, and I hope you can segue into it, and I repeat it if you forget it, is what role leadership?

I mean, you at HPE have been pushing this both internally to your company and now externally. How are you not only looking at it from the sort of avoiding compliance internally as we build this, but really making sure that the world embraces this as an opportunity for impact, not just another tool that kind of helps us carry on?

John Schultz

So let me take those two and combine it with the question that you asked the others, which is so my expectation is we can materially reduce, if not entirely eradicate, forced labor from the corporate economy.

What I mean by the corporate economy is where you have corporate supply chains and where traditional organizations and governments are engaged. When you’re into the black market and things like that, right, that’s going to be a different level of things. But the reason I say that is goes back to even this notion around shame and the like.

With the exception of the people who make money directly off of the slave trade, no one is pro-slavery. This is not a controversial issue, right? And again, not to open a can of worms with climate, but you’ve got people who are like, well, it’s not junk science and all that.

We don’t have any of that in slavery. There’s nobody who’s like, look, let’s talk slavery. I mean, there’s some positive elements to it, right?

So the only thing that allows it to thrive is what Amy and Mahendra have been saying, which is darkness, the information gap, right? So we have to turn the model around where companies and governments and the like, they’re not ashamed of saying, I found something. They’re ashamed of not looking and allowing the information gap to continue.

And in reality, I think in a lot of companies, it’s not that they’re looking the other way. I think they’re doing the best they reasonably can with the resources they have and the fact that it’s not a top priority. And you can feel good about what you’re doing, and the reality is you’re not making a damn bit of difference.

And that’s the problem we have, right? So, that’s why I’m so excited about the partnership. That’s why I think we can have this impact, because there are lots of causes to slavery and forced labor.

And we can’t address all of them. Economic disparity, et cetera, et cetera, across the globe, et cetera. But the one common ingredient that we can go attack is their ability to hide in the information gap and in the darkness.

This is truly one where sunlight is the best disinfectant. It is the total disinfectant. And that’s what we really need to do.

And so, we have to change the mindset at the corporate level, government level. Don’t be ashamed of what you’re finding. Be ashamed of the fact that you’re not looking.

Be ashamed that you’re not part of the data partnership. And therefore, you’re creating darkness and an information gap that the bad people are exploiting. Be part of the solution, and we will achieve this expectation.

It’s all right there. All right there in front of us. And we don’t have to guess at it, because the people who are living it every day, their organizations are doing this amazing work, are telling you, just don’t let these people hide any longer, so let’s go find them.

Dan Viederman

Beautifully said. I think, right, be ashamed that you’re not part of the data partnership is one of the messages. John.

I mean, we’re open, right? We’ve got a lot of really important partners around the table, from the private sector, from the intergovernmental sector, the UN system, from governments as well, in the form of the Ministry of Foreign Affairs from Thailand.

I think, from my perspective, as someone who’s been in the NGO side, and now on this sort of investment side, but has known many other people who have been working on this for a long time, we have no choice but to explore alternatives to the current system.

There is a wealth of solutions out there. We need to pull them together, and one way to pull them together is to illuminate them. Where are they bringing data to the table?

Where are they helping us understand the problem? But very importantly also, where can they help us understand what works and what doesn’t? It’s not simply a matter of identifying.

It’s also a matter of then remediating. ideally then leading in the direction of preventing, taking policy steps or changing sourcing practices or engaging workers at scale in a different way, so that the problem doesn’t emerge anymore. That’s ultimately the challenge in front of us.

I think the data partnership as it’s structured right now gives us a really strong chance to make progress. Ultimately, we should hold ourselves accountable, those of us who are part of it, as to whether or not it leads in the direction of measurable reduction in incidence of forced labor, both in supply chains, in civil society, as governments experience it as well.

With that, I will thank you all very much for being part of this panel and so powerfully talking about this issue that doesn’t get as much attention as it probably should for its importance to individuals, governments, and also businesses.

Thanks to those of you in the audience in person and those of you online. I’m grateful for your participation. If you want to learn more, go to the website, send us a note, join, be part, be willing to share your insight, share your data, share your tools.

We’re in open process right now. Thank you, everybody. Thank you.

Thank you. Thank you. Thank you so much.

John Schultz

Speech speed

168 words per minute

Speech length

Disagreed with

– Mahendra Pandey

Disagreed on

Scope of achievable impact

Sunlight and transparency are the best disinfectants against forced labor

Explanation

Schultz argues that transparency and information sharing are the most effective tools for combating forced labor. He emphasizes that eliminating the ability of bad actors to hide in darkness is the key to solving the problem.

Evidence

Recognition that while there are many causes of forced labor that cannot be addressed, the common ingredient of information gaps can be directly attacked

Major discussion point

Systemic Change and Future Vision

Topics

Human rights | Legal and regulatory

Kara Hurst

Speech speed

173 words per minute

Speech length

1534 words

Speech time

531 seconds

Analogy to carbon work where actual inputs versus modeled data enables faster decarbonization decisions with business leaders

Major discussion point

Worker Engagement and Empowerment

Topics

Human rights | Development

Amy Pope

Speech speed

166 words per minute

Speech length

845 words

Speech time

304 seconds

Cross-sectoral collaboration with protected data can transform detection capabilities

Explanation

Pope describes the current initiative as a ‘game-changer’ that will fundamentally remove the shield traffickers have used to hide their activities. She emphasizes that the ability to safeguard data while sharing it, combined with migrant worker input and government partnership, creates unprecedented detection capabilities.

Evidence

Agreed with

– Mahendra Pandey

Agreed on

Worker engagement and education are fundamental to effective solutions

Prevention through border management and worker education is as important as detection

Explanation

Pope emphasizes that the data partnership should enable both remedial and preventative approaches, including better border management, pre-migration education, and empowering workers to report violations with government capacity to respond.

Evidence

IOM’s successful impact from educating migrant workers about rights and providing knowledge about where to go and who to call when problems arise

Major discussion point

Implementation and Accountability

Topics

Human rights | Legal and regulatory

25 years of progress moved from denial to recognition, now technology enables action

Explanation

Pope provides historical context showing how the field has evolved from people not believing modern slavery exists to universal recognition of the problem, with current technology finally enabling effective action against increasingly sophisticated trafficking methods.

Evidence

Personal experience working on the issue for 25 years, from initial disbelief to current recognition but previous inability to detect due to supply chain complexity and trafficker sophistication

Major discussion point

Systemic Change and Future Vision

Topics

Human rights | Development

Mahendra Pandey

Speech speed

168 words per minute

Speech length

1709 words

Speech time

607 seconds

Worker engagement in data partnership design builds trust and ensures ground truth

Explanation

Pandey argues that without worker engagement, it’s impossible to understand the real problems or know whether solutions will work. He emphasizes that worker participation in the design process makes them responsible for trusting and implementing the system while sharing authentic information.

Evidence

Disagreed with

– John Schultz

Disagreed on

Emphasis on technology versus human engagement in solution design

Trust building with workers requires privacy protection and preventing retaliation

Explanation

Pandey explains that workers will only share authentic data if there is trust, which requires ensuring their privacy is secured and they won’t face deportation or job loss for sharing their experiences. This trust-building is essential for the system to work effectively.

Evidence

Recognition that workers face risks of deportation and job loss when reporting problems, requiring careful protection measures

Major discussion point

Worker Engagement and Empowerment

Topics

Human rights | Legal and regulatory

Real worker stories provide authentic data that audit findings cannot capture

Explanation

Pandey argues that company audits and UN reports cannot provide the real data about forced labor conditions. He contends that authentic information about recruitment fee abuses, deportations, and living conditions can only come from engaging directly with workers.

Evidence

Disagreed with

– John Schultz

Disagreed on

Scope of achievable impact

Dan Viederman

Speech speed

189 words per minute

Speech length

1905 words

Speech time

603 seconds

Measurable reduction in forced labor incidence should be the ultimate accountability metric

Explanation

Viederman establishes that the ultimate measure of success for the data partnership should be whether it leads to measurable reduction in the incidence of forced labor across supply chains, civil society, and government contexts. He emphasizes the importance of holding participants accountable to this outcome-focused standard.

Evidence

Recognition that there are many existing solutions that need to be pulled together and illuminated, with focus on both remediation and prevention

Major discussion point

Systemic Change and Future Vision

Topics

Human rights | Development

Agreements

Agreement points

Data sharing and cross-sectoral collaboration are essential for combating forced labor

Speakers

– John Schultz
– Kara Hurst
– Amy Pope
– Mahendra Pandey

Arguments

Data sharing enables more effective outcomes through insights generation

Sharing aggregated data across industries accelerates learning and risk identification

Cross-sectoral collaboration with protected data can transform detection capabilities

Worker engagement in data partnership design builds trust and ensures ground truth

Summary

All speakers agree that bringing together disparate data sources from various stakeholders (survivors, governments, NGOs, companies) through collaborative partnerships is crucial for making progress against forced labor. They emphasize that isolated efforts are insufficient and that shared insights lead to more effective outcomes.

Topics

Human rights | Development | Economic

Technology should augment human decision-making rather than replace it

Speakers

– Kara Hurst
– Mahendra Pandey

Arguments

Technology should augment human decision-making, not replace it

Worker engagement in data partnership design builds trust and ensures ground truth

Summary

Both speakers emphasize the critical importance of maintaining human oversight and engagement in AI-powered systems. They agree that technology should provide insights and analysis, but humans must remain central to decision-making processes and system design.

Topics

Human rights | Development

Worker engagement and education are fundamental to effective solutions

Speakers

– Amy Pope
– Mahendra Pandey

Arguments

Worker education before migration and capacity to report violations are crucial

Worker engagement in data partnership design builds trust and ensures ground truth

Real worker stories provide authentic data that audit findings cannot capture

Summary

Both speakers strongly advocate for meaningful worker engagement, emphasizing that workers must be educated about their rights and actively involved in solution design. They agree that without worker participation, systems cannot capture ground reality or build necessary trust.

Topics

Human rights | Development

Information gaps and lack of transparency enable forced labor to persist

Speakers

– John Schultz
– Amy Pope

Arguments

Information gaps and darkness allow forced labor to thrive in corporate supply chains

Cross-sectoral collaboration with protected data can transform detection capabilities

Summary

Both speakers identify the fundamental problem as traffickers and bad actors hiding in information gaps and darkness. They agree that eliminating these gaps through transparency and data sharing is essential for combating forced labor effectively.

Topics

Human rights | Economic

Human rights | Legal and regulatory

Unexpected consensus

Arguments

Agentic AI allows extracting insights without transferring sensitive data

Without worker engagement, AI systems cannot capture ground reality

Summary

Schultz emphasizes the technological solution of agentic AI as the ‘secret sauce’ that enables data sharing without compromising confidentiality, while Pandey stresses that technology without worker engagement in the design process cannot capture ground truth or ensure the system will work effectively.

Topics

Human rights | Development

Scope of achievable impact

Speakers

– John Schultz
– Mahendra Pandey

Arguments

Material reduction or eradication of forced labor from corporate economy is achievable

Complete prosperity requires addressing worker exploitation and trauma

Summary

Schultz expresses confidence that forced labor can be materially reduced or eradicated from corporate supply chains, focusing on corporate contexts, while Pandey argues for a broader systemic change that addresses the complete spectrum of worker exploitation and trauma as necessary for true prosperity.

Topics

Human rights | Development

Unexpected differences

Resource allocation priorities

Speakers

– Mahendra Pandey

Arguments

Investment in migrant justice deserves resources comparable to other development projects

Explanation

Pandey uniquely challenges the broader resource allocation priorities of society, questioning why entertainment infrastructure receives significant investment while migrant justice does not. This represents an unexpected critique of societal priorities that other speakers did not address, suggesting a more fundamental disagreement about how society values different types of development.

Topics

Human rights | Development

Overall assessment

Summary

The discussion shows remarkably high consensus on the core problem and general solution approach, with disagreements primarily centered on emphasis and scope rather than fundamental opposition. The main tensions are between technology-focused versus human-centered approaches, and between targeted corporate solutions versus broader systemic change.

Disagreement level

Low to moderate disagreement level. The speakers are largely aligned on the need for the Global Data Partnership and its potential effectiveness. Disagreements are more about implementation emphasis and scope of ambition rather than fundamental conflicts. This suggests strong potential for collaborative progress, though attention to worker engagement and broader systemic issues may be needed to maintain coalition unity.

Partial agreements

Human rights | Legal and regulatory

Takeaways

Key takeaways

The Global Data Partnership Against Forced Labor represents a paradigm shift from traditional compliance approaches to proactive, technology-enabled detection and prevention

Agentic AI enables cross-sectoral data sharing while preserving confidentiality, allowing insights extraction without compromising sensitive information

Worker engagement is essential for system effectiveness – without migrant worker participation, AI systems cannot capture ground truth or build necessary trust

The initiative has moved from proof-of-concept to MVP stage with Thailand as the pilot country for full production implementation

Success requires shifting corporate mindset from shame about finding forced labor to shame about not actively looking for it

Technology should augment human decision-making rather than replace it, with humans remaining in the loop for all critical decisions

The corporate economy can achieve material reduction or complete eradication of forced labor through eliminating information gaps and darkness that enable exploitation

Prevention through worker education, border management, and policy changes is equally important as detection and remediation

Resolutions and action items

Move the Global Data Partnership into full production in Thailand during year two as proof-of-concept

Demonstrate measurable impact in Thailand to enable scaling to additional countries and partners

Continue building trust with migrant worker networks through privacy protection and preventing retaliation

Expand data sharing partnerships across sectors to improve predictive capabilities

Develop policy advocacy initiatives based on aggregated data insights

Increase investment in worker education and capacity building programs

Hold the partnership accountable for measurable reduction in forced labor incidence as the ultimate success metric

Unresolved issues

Specific technical implementation details for maintaining data privacy while enabling effective sharing

Mechanisms for ensuring worker safety and preventing retaliation when they report violations

Scalability challenges for expanding beyond Thailand to multiple countries simultaneously

Resource allocation and funding models for sustained global implementation

Integration with existing regulatory frameworks and compliance systems across different jurisdictions

Measurement methodologies for tracking actual reduction in forced labor incidence

Addressing forced labor in black market economies beyond corporate supply chains

Suggested compromises

Using agentic AI to balance data sharing benefits with confidentiality concerns by extracting insights without transferring raw data

Implementing human-in-the-loop decision making to address concerns about AI replacing human judgment

Starting with single-country implementation (Thailand) rather than attempting immediate global rollout

This comment is deeply moving and insightful because it transforms the power dynamic from one of extraction (getting data from workers) to one of mutual engagement and relationship-building. The invitation to ‘see our life’ challenges the distance between decision-makers and affected communities.

Impact

This personal invitation created an emotional climax in the discussion and reinforced the theme of genuine partnership versus tokenism. It challenged other panelists and the audience to move beyond technical solutions toward authentic human connection, influencing the moderator’s closing call for participation and engagement.

Overall assessment

Explanation

This questions the allocation of resources and prioritization of migrant worker protection in development agendas

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.