22 – 29 May 2026
HIGHLIGHT OF THE WEEK
Magninifica humanitas: Human dignity in the age of AI
Pope Leo XIV has issued Magnifica Humanitas, a groundbreaking encyclical outlining the Catholic Church’s approach to AI.
SIdenote. An encylical pastoral letter is a circular letter written by the pope for the whole Roman Catholic Church on matters of doctrine, morals, or discipline. It is also aimed beyond the Church, often shaping global ethical and policy debates on major issues like inequality, war, the environment, and technology.
When critics objected to Pope Leo XIII publishing Rerum Novarum on society, the economy and politics, the Pope noted that the proclamation of the Gospel cannot overlook the concrete lives of people. This is the same logic behind Magnifica Humanitas.
Anchored in millennia of history, Magnifica Humanitas provides an important antidote to the AI discourse framed as inevitability: the race cannot be stopped, the market cannot be slowed, geopolitical competition cannot be escaped, and humans must simply follow.
The encyclical rejects this inevitability. It insists that technological futures should be shaped by our choices.
The encyclical frames AI governance through two contrasting metaphors: the Tower of Babel and the walls of New Jerusalem. Babel represents technological ambition without humility, reaching upward through ambition, control, and self-exaltation, only to produce fragmentation, confusion, and collapse. The Jerusalem wall metaphor offers a different vision. It is built piece by piece, through cooperation, solidarity, dialogue, and responsibility. Furthermore, it requires diplomacy rather than domination, patience rather than acceleration, and the common good rather than the race for supremacy.
Ultimately, Magnifica Humanitas can be boiled down to a choice between two metaphors: Babel and Jerusalem. The encyclical warns us about sleepwalking into Babel and offers guidance to reaching Jerusalem by drawing on the cognitive, emotional, and spiritual strengths that define us as human beings.
How to fix the ‘operating system’ of humanity. The encyclical frames humanity’s “operating system” as rooted in Axial Age traditions that defined human moral, rational, and spiritual identity, later refined through Enlightenment-era thought. It warns that transhumanism and posthumanism, driven by AI and bioengineering, risk recasting humans as entities to be optimised rather than inherently dignified beings. Magnifica Humanitas focuses on this key dilemma, removing all trappings of techno, business, and other narratives that dominate our communication space. It also calls to revisit the core of humanity’s ‘operating system’ centred around the life and dignity of humans, arguing that humans should not be ‘optimised’ in the same way as technologies.
The encyclical warns about the danger of regarding AI systems as neutral and objective, noting that technology takes on the characteristics of those who devise, finance, regulate and use it.
Clarity on AI governance and accountability. On AI governance, the encyclical offers clarity. It challenges the technocratic paradigm, in which governance is treated as an expert exercise focused mainly on standards, risk management, and efficiency. Instead, the encyclical argues that AI governance must begin with human dignity and the common good. Decisions about AI should be made by the people and communities directly affected. Communities should not merely be consulted after rules and platforms have been set elsewhere. They must be able to shape, challenge, and correct the systems that impact their lives. This resonates strongly with the multistakeholder tradition in internet governance, and challenges both corporate monopoly and state-centric control.
Fresh look at human rights. The encyclical points to a paradox of our age: human rights are widely formally proclaimed, yet rapidly eroded to, among others, by technological progress. Through profiling, manipulating and overall technological optimising, people are increasingly treated as data points and parts of algorithmic patterns. Therefore, one of the encyclical’s most critical contributions is its renewed grounding of human rights to protect our core humanity in the AI era.
Disarming AI. A striking concept in the encyclical is the call to ‘disarm AI.’ This is not limited to the regulation of autonomous weapons or military systems, although those are urgent concerns. It is a wider call for cognitive, economic, and political disarmament. AI is becoming an instrument in geopolitical rivalry, commercial dominance, surveillance, propaganda, and social control. The encyclical challenges the assumption that the AI race is inevitable. It calls for AI to be developed as an enabling tool for humanity rather than as a weapon of power and war.
Slavery and colonialism in the AI era. Digital colonialism is another concern of the encyclical. Colonial powers no longer need to control territory. They can appropriate data, shape markets, control infrastructures, and extract value from human lives transformed into exploitable information. A genuinely post-colonial digital order would return agency to individuals and communities. It would give people not only access to their data but meaningful control over how that data is used, by whom, and for whose benefit.
How to deal with AI risks. Magnifica Humanitas provides an informed and impactful survey of AI risks. On existential risk, it departs from narratives centred on AGI or super-intelligence takeover, arguing instead that existential risks may emerge more gradually, and not necessarily from technology itself, but from those who own and operate AI systems. On exclusion risk, it warns against concentration of power in the hands of a few and exclusion of many, with monopolies taking epistemic, economic, and political forms. It highlights how opacity and lack of public oversight can produce dependencies, exclusions, manipulations, and inequalities, including through control over visibility, amplification, and behavioural influence. It also examines existing risks in jobs, the economy, disinformation, and education.
Language and culture in the AI era. The Pope warns that violence often begins with language: wars are preceded by wars of words. Here, AI plays the critical role in shaping culture as a generator of text, video and other communication artefacts. The encyclical, therefore, calls for renewed attention to truth, dialogue, and the moral ecology of language. AI poses the main risks of generating synthetic narratives, shallow persuasion, and automated distortion. The challenge is not only misinformation. It is the gradual erosion of shared reality.
Education in the age of immediacy. Magnifica Humanitas places great emphasis on education in the AI era, in line with the Catholic Church’s long teaching tradition. The encyclical identifies a key tension between the rapid access to information in the digital age and the inherently slow nature of learning. Information can be delivered instantly. Knowledge and even more wisdom require a lot of time. However, schools and universities are not prepared to address the tension between AI technology and human learning.
Impact of AI on nature and the environment. The encyclical also reminds us that nothing in the digital world is immaterial or magical. AI depends on physical infrastructures: data centres, energy grids, minerals, water, devices, cables, and supply chains. Current AI systems require enormous amounts of energy and water and place growing pressure on natural resources. The environmental cost of AI is therefore not external to the debate. A technology that claims to advance humanity cannot be evaluated only by speed, accuracy, or profitability. It must also be judged by its impact on the planet, on vulnerable communities, and on future generations.
Why Magnifica Humanitas matters. The historic importance of Magnifica Humanitas lies in its attempt to shift the AI debate from technical management to civilisational reflection.
It challenges the concentration of power in a few companies. It questions transhumanist and posthumanist narratives that treat the human person as obsolete or upgradeable. It insists that AI governance must be rooted in dignity, subsidiarity, solidarity, justice, and the common good. It provides new perspectives on the risks of digital slavery, colonialism, militarisation, manipulation, and anthropological regression.
Most importantly, it restores moral agency. Metaphorically speaking, we are not condemned to build a Tower of Babel. We can choose to build the walls of the new Jerusalem.
This text is an adaptation of Dr Jovan Kurbalija’s analysis of Magnifica Humanitas, which will be expanded in the coming days. To receive in-depth analysis, you can register here.
IN OTHER NEWS LAST WEEK
AI governance
China. ByteDance has reportedly agreed to purchase millions of Qualcomm ASIC chips for use in its expanding AI infrastructure and agent systems. The deal emerged shortly after President Trump’s visit to China and reflects continued interdependence between US semiconductor firms and Chinese AI platforms despite broader export-control tensions.
North Korea. Pyongyang has announced the successful testing of AI-guided cruise missiles and precision artillery systems, signalling a new stage in the country’s efforts to modernise its conventional military capabilities. State media said the systems incorporate AI-assisted targeting and autonomous navigation technologies designed for modern warfare.
Tech CEOs. Public remarks by senior figures at Anthropic and OpenAI have exposed widening divisions within the AI sector over how rapidly advanced systems will reshape white-collar employment. Speaking at a Vatican conference on AI ethics, Anthropic co-founder Chris Olah warned that large-scale labour displacement remains a credible outcome as frontier models become more capable. OpenAI CEO Sam Altman has meanwhile adopted a more optimistic position, arguing that AI has so far had a smaller-than-expected impact on entry-level white-collar employment and suggesting earlier predictions of rapid job elimination may have overstated the pace of disruption. The contrasting messaging reflects growing divergence among leading AI labs over how to frame the economic implications of frontier AI, complicating efforts by governments, businesses and labour markets to assess the scale and timing of workforce transformation.
Kazakhstan. The UN Educational, Scientific and Cultural Organization (UNESCO) has announced the launch of its AI Readiness Assessment Methodology in Kazakhstan to evaluate the country’s preparedness for AI governance and development. The framework is intended to help countries align AI governance approaches with the UNESCO Recommendation on the Ethics of AI.
Argentina. Argentina’s Ministry of Human Capital has launched the ‘Digital Twin’ initiative, an AI-based system intended to simulate potential impacts of social policies before implementation. The system is designed to model scenarios related to areas including poverty, subsidies, and human capital development using large-scale datasets. Officials said the project is part of broader efforts to use data analysis and predictive tools in public policy planning.
Greece. Greece’s Ministry of Digital Governance and Artificial Intelligence has launched ‘Artificial Intelligence for All’, a public guide designed to improve understanding and use of AI tools. The material introduces basic concepts related to AI and large language models through practical examples and simplified explanations. The guide explains how AI systems can process different forms of data and generate outputs, including recommendations, summaries, and digital content.
Spain. Spain’s Council of Ministers has approved a draft Organic Law aimed at adapting the EU AI Act into the country’s national legal framework. The draft law designates Spain’s Artificial Intelligence Supervisory Agency, based in A Coruña, as the central authority. Other market surveillance authorities will also have roles, including the Bank of Spain for financial systems, the Spanish Data Protection Agency for data-related matters, and the General Council of the Judiciary for justice-related issues.
Vietnam. Vietnam will require disclosure labels for certain AI-generated and AI-edited content under a new government decree aimed at improving online transparency, starting from May 2026. Under Decree 142/2026/ND-CP, organisations and individuals using AI systems must disclose when content has been created or altered by AI in ways that could affect perceptions of authenticity. The rules apply to AI-generated or AI-edited audio, image, and video content, particularly content that imitates real people or realistic events. Particularly, it applies to content that imitates the appearance or voice of real people or recreates real-life events in a convincing manner. According to the decree, disclosures must be clear, visible, and recognisable before or during user access to the content.
South Africa. The Department of Communications and Digital Technologies of South Africa has established an independent seven-member expert panel to review a draft national AI policy, following the earlier version’s withdrawal due to fictitious and potentially AI-generated references. The new panel has been mandated to revise or remove problematic content and correct citations, with a revised policy expected to reach Cabinet approval by November 2026 and public consultation targeted for January 2027.
Mythos. Mythos. Anthropic has outlined the next phase of its Project Glasswing cybersecurity programme. The programme will be expanded to additional government and allied partners, with further tooling and research support intended to strengthen broader organisational cyber resilience. The company adds that it ultimately plans to release Mythos-class models more widely once stronger safety mechanisms are developed, framing the longer-term goal as a shift toward a more secure software.
Europe’s tech sovereignty push and pull
The European Commission has once again delayed its long-discussed tech sovereignty package, reportedly for the third time. The package is expected to include measures to reduce dependence on non-European providers in strategic sectors such as cloud services and AI infrastructure. Reports suggest concerns from the USA over potentially protectionist measures may have contributed to the latest delay, with the package now scheduled to be presented on 3 July.
The delay comes as Europe debates a broader question: can it build digital autonomy while remaining deeply dependent on American technology companies?
That tension is now visible across multiple policy fronts.
One of the clearest examples emerged in the Netherlands, where concerns over sovereignty have reached the level of national digital identity infrastructure. Dutch citizens and privacy advocates challenged the proposed acquisition of Solvinity, the company hosting the Netherlands’ DigiD identity system, by US-based Kyndryl. Critics warned that allowing an American company near such sensitive infrastructure could expose Dutch citizens’ data to foreign jurisdiction and surveillance risks under US laws.
Earlier in May, the District Court of The Hague rejected an attempt by three Dutch citizens to block the government from renewing its contract with Solvinity. However, the Investment Assessment Office (BTI) has advised Willemijn Aerdts, Minister for the Digital Economy and Sovereignty of the Netherlands, to impose a complete ban on this acquisition, as it may pose a risk to the public interest. The minister has acted upon this advice, and the government has blocked the sale.
Meanwhile, Brussels is also moving to reserve more strategic telecom and satellite infrastructure for European operators.
This week, the Commission adopted a proposal for the future allocation of the EU’s 2 GHz mobile satellite services spectrum band beyond 2027. The spectrum is increasingly viewed as a strategic asset for both commercial connectivity and defence-related communications, particularly for direct-to-device (D2D) satellite services and critical communications infrastructure.
Under the proposal, one-third of the spectrum would be reserved for governmental, security, and military use, tied to the EU’s IRIS² secure connectivity programme. The remaining two-thirds would support commercial services, but half of that capacity would be reserved specifically for EU operators entering the market, while the rest would remain open to both EU and non-EU companies.
The proposal reflects Europe’s growing efforts to secure control over critical digital infrastructure while maintaining some openness to foreign providers.
The same debate is increasingly visible in defence AI. A year after NATO adopted Palantir’s AI-enabled battlefield system, concerns are growing over Europe’s lack of homegrown alternatives and continued dependence on US defence technology companies. French AI company Mistral this week defended its military AI work, arguing Europe needs its own capabilities rather than relying entirely on foreign providers.
The big picture. Europe’s real test ahead is deciding how much dependency is acceptable in each layer of its digital stack. Mark your calendars for next Wednesday, though, as the tech sovereignty package should let us know more about the bloc’s plans.
EU adopts unified cyber incident reporting templates under NIS2
The EU has adopted common templates for cybersecurity incident reporting across the bloc, marking a step towards more harmonised compliance requirements under the NIS2 Directive.
The templates were adopted during the 39th plenary meeting of the NIS Cooperation Group, comprising EU member states, the European Commission, and the EU Agency for Cybersecurity (ENISA).
According to the Commission, the standardised templates are designed to reduce administrative burdens and simplify compliance for companies required to report cybersecurity incidents under NIS2. The move also aligns with broader EU efforts to create a single-entry point for incident reporting under the proposed Digital Omnibus initiative.
What’s next? The Commission now plans to adopt the templates through an implementing act, which would make them mandatory for all member states.
Iran restores internet access after 88-day nationwide shutdown
After 88 days of nationwide shutdown, internet connectivity has been partially restored in Iran, ending one of the longest full-scale blackouts in modern history.
The shutdown was imposed after escalating conflict dynamics and justified by authorities on security grounds.
On Tuesday, First Vice-President Mohammad Reza Aref wrote on X that ‘The first step toward free and regulated access to cyberspace has been taken.’
NetBlocks reported that connectivity has largely returned, but that metrics indicate that users still face heavy filtering, including new restrictions on messaging and app stores.
What’s next? NetBlocks’ research director, Isik Mater, told BBC: ‘Historically, each time internet access has been restored after an internet shutdown in Iran it has come back with heavier restrictions and tighter controls.’
Meta faces ongoing legal pressure as Supreme Court declines to hear Vermont addiction case
The US Supreme Court denied a petition for certiorari, a discretionary mechanism by which a higher court decides whether to review a lower court’s decision, in a social media addiction case brought by the state of Vermont against Meta Platforms Inc.
Vermont accused Meta, the parent company of Facebook and Instagram, of knowingly designing addictive platform features that harm minors and negatively affect their mental health.
Meta argued that Vermont courts lacked jurisdiction due to insufficient contacts with the state. However, by denying certiorari, the Supreme Court left the Vermont Supreme Court’s decision in place.
Vermont Attorney General Charity Clark welcomed the decision, stating that it affirms ‘that companies that choose to do business in Vermont, like Meta, can be held accountable when they harm kids’.
Why does it matter? The case is part of ongoing litigation across the USA against social media companies. The litigation could have major implications for the future legal responsibility of technology companies toward minors and online platform safety.
EU fines Temu for DSA breach
The European Commission issued a fine of €200 million to Temu after finding that Temu’s 2024 risk assessment fell short of the standards laid out in the Digital Services Act (DSA). It relied on general information about risks concerning the e-commerce sector, rather than Temu-specific evidence. It underestimated the likelihood that EU consumers would encounter illegal or unsafe products. The Commission further found that Temu did not adequately assess how its platform design, particularly recommender systems and influencer-driven promotion, could contribute to the amplification and spread of illegal products.
The fine issued was calculated taking into account the nature of the infringement, its gravity in terms of affected EU users, and its duration.
What’s next? Following the decision, Temu is required to submit an action plan to the Commission by 28 August 2026 outlining measures to correct its risk assessment deficiencies, in accordance with Article 75 of the DSA. The European Board for Digital Services will review the plan and issue an opinion within one month, after which the Commission will have a further month to adopt a final decision and set a deadline for implementation.
If Temu fails to comply, the Commission may impose periodic penalty payments. The Commission continues to monitor and ensure compliance with both the decision and the DSA obligations.
Georgia prepares launch of GEL₮ stablecoin tied to national currency
Georgia is preparing to launch GEL₮, a stablecoin linked to the Georgian Lari. According to officials, the initiative is intended to support blockchain-based payment infrastructure under a dedicated stablecoin regulatory framework.
Officials said the project could support faster payments, cross-border transfers, fintech development, and programmable financial services.
The project builds on Georgia’s digital asset frameworks, which include measures related to reserve management, redemption rights, issuer oversight, and anti-money laundering compliance.
Why does it matter? Georgia’s decision to support a national currency stablecoin reflects a broader shift towards blockchain-based settlement systems that could reshape cross-border payments, remittances, and digital commerce.
Papua New Guinea advances digital ID law
Papua New Guinea’s Ministry of Information and Communications Technology has begun drafting instructions and a proposed bill for digital identity and verifiable credentials legislation following the endorsement of the National Digital ID Policy.
ICT Minister Peter Tsiamalili Jr said the process marks a step towards a legal framework that will allow citizens to identify themselves securely and access trusted digital services.
The proposed legislation will support the national rollout of SevisPass, SevisWallet, SevisDEx, and other approved verifiable credentials. SevisWallet will allow citizens to register, hold, and present trusted digital credentials, while SevisDEx will enable secure, consent-based data exchange.
What’s next? The ministry said relevant agencies, issuers, verifiers, and relying partners should align their systems and compliance pathways to support the rollout by July 2026.
LAST WEEK IN GENEVA
UN Human Rights Office issues guidelines on child safety online
A new brief by the Office of the UN High Commissioner for Human Rights (OHCHR) outlined 10 interconnected steps to improve children’s online safety through a rights-based, system-level regulatory approach. Rather than relying on broad restrictions or shifting responsibility to users, it prioritises structural changes to how platforms are designed and governed.
First, it calls for safety-by-design obligations that reduce harmful features such as autoplay, infinite scroll, and intrusive notifications. Regulation should be grounded in children’s rights, ensuring that protection is balanced with rights to participation, expression, privacy, and access to information. Strong data privacy protections are central, including prohibiting micro-targeting and limiting the profiling of children for commercial purposes.
Companies are expected to carry out ongoing child rights impact assessments as part of broader human rights due diligence, with transparency and accountability for their findings. Where age verification is used, it must include strict safeguards around data collection, storage, and non-discrimination.
The guidance emphasises targeted, harm-specific age restrictions rather than blanket bans, and stresses that children should be actively involved in shaping the policies that affect their digital environments. Platforms should also be required to provide meaningful transparency on recommender systems, moderation practices, and design choices.
Finally, effective oversight, enforceable accountability mechanisms, and continuous, evidence-based policymaking are essential to ensure that regulation adapts as technologies and risks evolve.
Notably, the guidance frames online child safety as both a governance issue and a rights issue. It argues that children’s rights apply online as fully as they do offline, including rights related to participation, privacy, access to information, and freedom of expression. Children, the brief notes, should also have a say in how digital spaces are governed and regulated.
LOOKING AHEAD
The WTO will organise the Trade and Environment Week 2026, a week-long series of meetings and thematic sessions held from 1–5 June 2026 at the WTO Headquarters in Geneva and online. The programme brings together WTO members and invited stakeholders to discuss the intersection of international trade and environmental policy. The 2026 edition includes sessions on topics such as eco-design and digital product passports, carbon accounting and emissions standards, trade in environmental goods and services, biodiversity, climate-related border measures, fossil fuel subsidy reform, and the role of trade in supporting global climate action.
The 114th International Labour Conference will convene in Geneva from 1–12 June 2026, bringing together governments, employers, and workers from ILO member states. The agenda includes standard-setting discussions on decent work in the platform economy, a general discussion on advancing gender equality in the world of work, and oversight of international labour standards implementation. Delegates will also consider programme, budget, and institutional matters, including supervisory reports and procedural questions. The Conference remains the ILO’s key forum for setting global labour standards and shaping policy responses to evolving labour market challenges.
Geneva in Action: Scaling School Connectivity Worldwide (Part II) will be held on 3 June at the Giga Connectivity Centre in Geneva. Co-hosted with Diplo, the 75-minute session brings together Geneva-based organisations and government officials participating in the Giga Learning Hub. The discussion will focus on scaling school connectivity within broader digital transformation and e-government reforms, with emphasis on national ownership, global frameworks, and the balance between digital sovereignty and locally adapted technology solutions. Participants will also examine opportunities and risks in delivering universal, meaningful connectivity for schools and communities through international collaboration and policy alignment across stakeholders.
READING CORNER
Pope Leo XIV’s first encyclical takes on the technocratic paradigm, transhumanism, and the TINA (There Is No Alternative) mindset of the AI era.
DON’T MISS
Follow the road to the 2027 Geneva AI Summit, a global journey shaping AI’s future. Bookmark the page and register for updates to stay informed and involved every step of the way to Geneva 2027.