Content Policy

Airbnb faces allegations of compromising user safety

Airbnb has been accused of compromising user safety by scaling back efforts to remove extremists from its platform, according to a whistle-blower complaint by Jess Hernandez, a former contractor. Hernandez, who worked as an investigations analyst for Airbnb from May 2022 to November 2023, claims she was fired after the company directed her team to reinstate users involved in the 6 January 2021 Capitol attack. Whistle-blower Aid, representing Hernandez, stated that Airbnb’s changes undermined its public safety commitment.

Hernandez filed her complaint with the US Securities and Exchange Commission and Federal Trade Commission in May. Airbnb denied the allegations, asserting that it continues to enforce policies against dangerous individuals and has even expanded its team to enhance safety measures.

Despite these measures, Hernandez alleges that in 2023, the teams faced increased bureaucratic hurdles, slowing down their ability to remove dangerous users. The claim is supported by a 161-page complaint obtained by NBC from an anonymous source. Before her time at Airbnb, Hernandez worked with the Terrorism Research and Analysis Consortium.

Why does it matter?

The complaint adds to ongoing safety concerns within Airbnb, a platform facilitating millions of global interactions. CEO Brian Chesky has previously implemented measures like party crackdowns and bans on indoor security cameras to address these issues. Airbnb’s history of removing users associated with extremist activities dates back to 2016, including actions following the Unite the Right rally in 2017 and the Capitol attack in 2021.

Tencent to remove game from some Android app stores

Tencent is set to remove its popular mobile game ‘Dungeon & Fighter’ (DnF Mobile) from selected Android app stores starting Thursday due to expired contracts. The Chinese tech giant did not specify which app stores will be affected, but local media reports indicate that Huawei, Oppo, and Vivo are among them.

Why does it matter?

The issue highlights ongoing tensions between game developers and distributors in China, particularly over the mobile game market’s standard 50% revenue share split. In 2021, Tencent faced a similar issue when Huawei removed several of its mobile games from its app store over revenue-sharing disagreements.

Meta and Google adopt different approaches to election-related query restrictions in India

As India’s elections conclude and the new government commences its term, Meta has removed restrictions on election-related queries through its Meta AI chatbot. Users can now access information about election results, politicians, and officeholders. Initially, Meta had limited such queries, directing users to the Election Commission’s website for information on politicians, candidates, and political parties. While Meta hasn’t issued an official statement, this move aligns with the company’s ongoing efforts to refine its AI models.

Meanwhile, despite launching its Gemini AI app in India, Google maintains restrictions on election-related queries as part of a global policy. The company directs users to Google Search instead of providing direct responses through Gemini AI. These restrictions were implemented earlier this year in response to elections worldwide. However, it remains to be seen when Google will lift these restrictions, particularly in countries where elections have concluded and new governments are in place.

Why does it matter?

The differing approaches of Meta and Google highlight the complexities surrounding AI chatbots and political queries. While Meta temporarily restricted queries during the Indian elections, Google maintains global restrictions. The decisions underscore companies’ challenges in managing AI outputs, especially amidst concerns about bias and misinformation. Other AI chatbots like ChatGPT and Microsoft Copilot also exhibit varied responses to political queries, reflecting the broader scrutiny developers face in ensuring the integrity of AI-driven platforms.

AI chatbot’s mayoral bid halted by legal and ethical concerns in Wyoming

Victor Miller, 42, has stirred controversy by filing to run for mayor of Cheyenne, Wyoming, using a customised AI chatbot named VIC (virtual integrated citizen). Miller argued that VIC, powered by OpenAI technology, could effectively make political decisions and govern the city. However, OpenAI quickly shut down Miller’s access to their tools for violating policies against AI use in political campaigning.

The emergence of AI in politics underscores ongoing debates about its responsible use as technology outpaces legal and regulatory frameworks. Wyoming Secretary of State Chuck Gray clarified that state law requires candidates to be ‘qualified electors,’ meaning VIC, as an AI bot, does not meet the criteria. Despite this setback, Miller intends to continue promoting VIC’s capabilities using his own ChatGPT account.

Meanwhile, similar AI-driven campaigns have surfaced globally, including in the UK, where another candidate utilises AI models for parliamentary campaigning. Critics, including experts like Jen Golbeck from the University of Maryland, caution that while AI can support decision-making and manage administrative tasks, ultimate governance decisions should remain human-led. Despite the attention these AI candidates attract, observers like David Karpf from George Washington University dismiss them as gimmicks, highlighting the serious nature of elections and the need for informed human leadership.

Miller remains optimistic about the potential for AI candidates to influence politics worldwide. Still, the current consensus suggests that AI’s role in governance should be limited to supportive functions rather than decision-making responsibilities.

New social network app blends AI personas with user interactions

Butterflies, a new social network where humans and AI interact, has launched publicly on iOS and Android after five months in beta. Founded by former Snap engineering manager Vu Tran, the app allows users to create AI personas, called Butterflies, that post, comment, and message like real users. Each Butterfly has unique backstories, opinions, and emotions, enhancing the interaction beyond typical AI chatbots.

Tran developed Butterflies to provide a more creative and substantial AI experience. Unlike other AI chatbots from companies like Meta and Snap, Butterflies aims to integrate AI personas into a traditional social media feed, where AI and human users can engage with each other’s content. The app’s beta phase attracted tens of thousands of users, with some spending hours creating and interacting with hundreds of AI personas.

Butterflies’ unique approach has led to diverse user interactions, from creating alternate universe personas to role-playing in popular fictional settings. Vu Tran believes the app offers a wholesome way to interact with AI, helping people form connections that might be difficult in traditional social settings due to social anxiety or other barriers.

Initially free, Butterflies may introduce a subscription model and brand interactions in the future. Backed by a $4.8 million seed round led by Coatue and other investors, Butterflies aims to expand its functionality and continue to offer a novel way for users to explore AI and social interaction.

US Justice Department to investigate TikTok over child privacy complaint

The US Federal Trade Commission (FTC) has referred a complaint against TikTok and its parent company, ByteDance, to the Justice Department over potential violations of children’s privacy. The move follows an investigation that suggested the companies might be breaking the law and deemed it in the public interest to proceed with the complaint. The following investigation stems from allegations that TikTok failed to comply with a 2019 agreement to safeguard children’s privacy.

TikTok has been discussing with the FTC for over a year to address the agency’s concerns. The company expressed disappointment over the FTC’s decision to pursue litigation rather than continue negotiations, arguing that many of the FTC’s allegations are outdated or incorrect. TikTok remains committed to resolving the issues and believes it has already addressed many concerns.

Separately, TikTok is facing scrutiny from US Congress regarding the potential misuse of data from its 170 million US users by the Chinese government, a claim TikTok denies. Additionally, TikTok is preparing to file a legal brief challenging a recent law that mandates its parent company, ByteDance, to divest TikTok’s US assets by 19 January or face a ban.

G7 Italy summit unveils AI action plan to balance AI risks and opportunities

Adopted on June 14, 2024, at the G7 Summit in Apulia, Italy, the Group of Seven (G7) Leaders’ Communiqué, expresses the wealthiest nations’ common pledges and actions to address multiple global issues. A portion of the Group of Seven (G7) declaration closing the Italian summit focuses on AI and other digital matters.

G7 leaders called for an action plan to manage AI’s risks and benefits, including developing and implementing an International Code of Conduct for organisations developing advanced AI systems, as unveiled last October under the Japanese G7 presidency. To maximise the advantages of AI while mitigating its threats, G7 nations commit to deepening their cooperation.

An action plan for the use of AI in the workplace was announced, together with the creation of a brand to promote the implementation and use of the International Code of Conduct for advanced AI systems, in cooperation with OECD. G7 leaders stressed the importance of global partnership to bridge the digital divide and ensure that people around the world have access to the benefits of AI and other technologies. The goal is to advance science, improve public health, accelerate the clean energy transition, promote sustainable development goals, etc.

Why does it matter?

The G7 is encouraging global collaboration within the group of countries, with the OECD, with other initiatives such as the Global Partnership on AI (GPAI), and towards the developing world, to facilitate the equitable distribution of the benefits of AI and other emerging technologies while minimising any threats. G7 leaders aim to mend technological gaps and address AI’s impact on workers. G7 labor ministers are tasked with designing measure to capitalize on AI’s potential, promote quality employment, and empower people, while also tackling potential barriers and risks to workers and labour markets.

G7 leaders agreed to intensify efforts to promote AI safety and enhance interoperability between diverse approaches to AI governance and risk management. That means strengthening collaboration between AI Safety Institutes in the US, UK, and equivalent bodies in other G7 nations and beyond, to improve global standards for AI development and implementation. The G7 also formed a ‘Semiconductors Point of Contact Group’ to strengthen cooperative efforts on addressing challenges affecting this critical industry that drives the AI ecosystem.

G7 nation’s commitments are consistent with the recent Seoul AI safety summit efforts and align with the intended goals of the upcoming United Nations Summit of the Future. Echoing the UN General Assembly landmark resolution on ‘seizing the opportunities of safe, secure, and trustworthy AI systems for sustainable development’ and Pope Francis’s historic address to the G7 leaders, the communiqué reflects the group’s unified stance on AI safety and the need for a framework for AI’s responsible development and use in the military.

Warning labels for social media suggested by US Surgeon General

US Surgeon General Vivek Murthy has called for a warning label on social media apps to highlight the harm these platforms can cause young people, particularly adolescents. In a New York Times op-ed, Murthy emphasised that while a warning label alone won’t make social media safe, it can raise awareness and influence behaviour, similar to tobacco warning labels. The proposal requires legislative approval from Congress. Social media platforms like Facebook, Instagram, TikTok, and Snapchat have faced longstanding criticism for their negative impact on youth, including shortened attention spans, negative body image, and vulnerability to online predators and bullies.

Murthy’s proposal comes amid increasing efforts by youth advocates and lawmakers to protect children from social media’s harmful effects. US senators grilled CEOs of major social media companies, accusing them of failing to protect young users from dangers such as sexual predators. States are also taking action; New York recently passed legislation requiring parental consent for users under 18 to access ‘addictive’ algorithmic content, and Florida has banned children under 14 from social media platforms while requiring parental consent for 14- and 15-year-olds.

Despite these growing concerns and legislative efforts, major social media companies have not publicly responded to Murthy’s call for warning labels. The push for such labels is part of broader initiatives to mitigate the mental health risks associated with social media use among adolescents, aiming to reduce issues like anxiety and depression linked to these platforms.

FCC names Royal Tiger as first official AI robocall scammer gang

The US Federal Communications Commission (FCC) has identified Royal Tiger as the first official AI robocall scammer gang, marking a milestone in efforts to combat sophisticated cyber fraud. Royal Tiger has used advanced techniques like AI voice cloning to impersonate government agencies and financial institutions, deceiving millions of Americans through robocall scams.

These scams involve automated systems that mimic legitimate entities to trick individuals into divulging sensitive information or making fraudulent payments. Despite the FCC’s actions, experts warn that AI-driven scams will likely increase, posing significant challenges in protecting consumers from evolving tactics such as caller ID spoofing and persuasive social engineering.

While the FCC’s move aims to raise awareness and disrupt criminal operations, individuals are urged to remain vigilant. Tips include scepticism towards unsolicited calls, utilisation of call-blocking services, and verification of caller identities by contacting official numbers directly. Avoiding sharing personal information over the phone without confirmation of legitimacy is crucial to mitigating the risks posed by these scams.

Why does it matter?

As technology continues to evolve, coordinated efforts between regulators, companies, and the public are essential in staying ahead of AI-enabled fraud and ensuring robust consumer protection measures are in place. Vigilance and proactive reporting of suspicious activities remain key in safeguarding against the growing threat of AI-driven scams.

AI tools struggle with election questions, raising voter confusion concerns

As the ‘year of global elections’ reaches its midpoint, AI chatbots and voice assistants are still struggling with basic election questions, risking voter confusion. The Washington Post found that Amazon’s Alexa often failed to correctly identify Joe Biden as the 2020 US presidential election winner, sometimes providing irrelevant or incorrect information. Similarly, Microsoft’s Copilot and Google’s Gemini refused to answer such questions, redirecting users to search engines instead.

Tech companies are increasingly investing in AI to provide definitive answers rather than lists of websites. This feature is particularly important as false claims about the 2020 election being stolen persist, even after multiple investigations found no fraud. Trump faced federal charges for attempting to overturn Biden’s victory, who won decisively with over 51% of the popular vote.

OpenAI’s ChatGPT and Apple’s Siri, however, correctly answered election questions. Seven months ago, Amazon claimed to have fixed Alexa’s inaccuracies, and recent tests showed Alexa correctly stating Biden won the 2020 election. Nonetheless, inconsistencies were spotted last week. Microsoft and Google, in return, said they avoid answering election-related questions to reduce risks and prevent misinformation,, a policy also applied in Europe due to a new law requiring safeguards against misinformation.

Why does it matter?

Tech companies are increasingly tasked with distinguishing fact from fiction as it develops AI-enabled assistants. Recently, Apple announced a partnership with OpenAI to enhance Siri with generative AI capabilities. Concurrently, Amazon is set to launch a new AI version of Alexa as a subscription service in September, although it remains unclear how it will handle election queries. An early prototype struggled with accuracy, and internal doubts about its readiness persist. The new AI assistants from Amazon and Apple aim to merge traditional voice commands with conversational capabilities, but experts warn this integration may pose new challenges.