Senator Murphy Demands AI Regulation as Nobel Laureate, Tech Leaders Sound Alarms at UN Forum

Foreign Policy's Tech and AI Forum during UN General Assembly Week brought together ministers, Nobel laureates, labor leaders, and tech executives for a sweeping debate on AI's threat to democracy, jobs, and global equity — with the sharpest call coming from a US senator who says Americans are being used as "guinea pigs" for unchecked technology.

• Senator Murphy calls for mandatory AI regulation within 12 months, including deepfake bans, friendship bot restrictions, and a third-party model approval process
• Nobel laureate Maria Ressa warns AI has "hacked" human attention, intimacy, and agency in three waves, with 74% of the world now under authoritarian rule
• McKinsey projects COVID-scale workforce disruption sustained for a decade, with 11 million US jobs in transition and 25 million requiring new skills
• Microsoft security chief says agentic AI has rewritten the "physics of cybersecurity", with 1 million AI agents running inside the company
• Mozilla Foundation argues open-source AI is a precondition for sovereignty, warning frontier model consolidation threatens global creativity
• AFL-CIO and faith leaders unite across partisan lines to demand worker voice in AI deployment, citing union-negotiated technology notice clauses
• GSMA reports 3.1 billion people have mobile coverage but remain offline, with a memory crisis doubling the cost of affordable smartphones

---

Murphy Says Inaction Makes Americans 'Guinea Pigs'

Why it matters: The most powerful US legislative voice on AI at the forum framed the domestic regulation debate in national security terms — arguing that China wants the US to remain unregulated.

Where things stand: Sen. Chris Murphy, Senator from Connecticut, laid out three "achievable" legislative steps: outlawing or watermarking deepfakes in political campaigns, banning friendship chatbots for children, and creating a third-party model approval process. He noted that 60% of US children already use chatbots for friendship and advice.

"Friendship bots, as I mentioned before, a poison for our kids, a poison, we could just outlaw those," said Sen. Murphy.

On the geopolitical framing that AI regulation would cede ground to China, Murphy flipped the argument: "China would love nothing more than for the United States to do nothing to protect people spiritually and the economy generally from what AI could do unchecked. They would love for us to be the one that submits our citizens as guinea pigs to this technology."

Murphy also drew a sharp line on industry influence, warning that companies should participate in regulatory design but never control it. "The last thing we should do is give them veto power. They will attempt to try to have veto power. They will try to spend money in elections to get folks in power who will only pass regulatory regimes that they approve."

The other side: Jeff Browning, Director of Global Government Affairs and Public Policy at ServiceNow, offered the industry perspective, advocating for AI kill switches that operate at machine speed, industry standards work through NIST, OECD, and ISO, and vendor accountability through long-term customer engagement — a framework that positions standards bodies rather than Congress as the primary governance mechanism.

What's next: Murphy expressed pessimism about the Trump-Xi summit yielding AI cooperation and signaled the American public is "begging on both sides of the aisle" for regulation — suggesting legislative action could advance regardless of the administration's posture.

---

Ressa: AI Has Already Hacked Your Biology

Why it matters: The Nobel Peace Prize laureate framed AI not as a future risk but as a decade-long assault on democratic foundations that has already succeeded in its first two phases and is now entering its most dangerous one.

Where things stand: Maria Ressa, Founder and CEO of Rappler and Co-chair of the UN Independent International Scientific Panel on AI, described three cascading phases: social media's attention economy (microtargeting and diminished attention spans), chatbots hacking intimacy (citing a 14-year-old who died by suicide after interacting with a Character AI chatbot), and agentic AI hacking human agency itself.

"We gave up our attention. We gave up our intimacy. Now what's being hacked is agency itself. These agents, literally, you give them permission to act for you," Ressa said.

She cited V-Dem data showing 74% of the world now lives under authoritarian rule and argued this correlates directly with social media's manipulation of emotions and voting behavior. She reported that 53% of internet traffic is now non-human and that Google's generative AI search has devastated journalism's remaining economic model.

"When you now search and generative AI gives you your answer without bringing you to our website, it means it took our content and did not bring you back, which means we have no revenue. The drop globally is around 75 to 80%," Ressa said.

Ressa discussed the UN panel she co-chairs with Turing Award winner Yoshua Bengio, which produced a consensus report in six weeks with 40 experts from 37 countries. She noted the US voted against creating the panel, with only Paraguay also opposing.

"Information integrity is the mother of all battles. We win that, we win everything. We lose that, we lose everything," she said.

---

COVID-Scale Job Disruption — for 10 Years Straight

Why it matters: The largest skills migration in American history is already underway, and half the workers who need to transition face significant barriers — especially in healthcare, the fastest-growing sector.

Where things stand: Tongyi Catlin, Director of McKinsey Global Institute, previewed forthcoming research projecting that 11 million US jobs will be in full transition while an additional 25 million jobs — 70% of the workforce — will retain their titles but require fundamentally different skills.

"Imagine what we've experienced during COVID. COVID was a massive, massive shock in the labor system in the US. Well, you take COVID and now we're going to need to sustain that level of magnitude and pace for 10 years," Catlin said.

Sectors like sales, office administration, and transportation are expected to decline, while healthcare will see the highest demand — but also the highest credentialing barriers. Only 1 in 7 transitioning workers will have an easy pathway; for 50%, the hurdles are significant.

Catlin argued the implications extend beyond workforce planning. As AI drives the cost of prediction toward zero, the value of human judgment rises correspondingly: "The cost of prediction is coming down to zero, which basically means that we're going to make predictions for everything, which basically means that the demand for judgment and the value of judgment is going to increase dramatically."

He cited Singapore's merger of its skills and workforce departments and Canada's self-assessment platform as models, and drew a historical parallel to farming, which represented 50% of US jobs in 1920 and just 3% fifty years later.

---

Agentic AI Rewrites Cybersecurity From the Inside

Why it matters: Defenses built for human threats are dangerously inadequate when AI agents — polymorphic, high-privileged, and operating at machine speed — become the new insiders.

Where things stand: Hayat Gallo, Executive Vice President of Security at Microsoft, identified three categories of threat: AI insiders (sanctioned or unsanctioned agents with high-privileged access that can be poisoned), AI-powered attack acceleration, and amplified exposure from existing vulnerabilities creating "10x exposure." Microsoft currently runs approximately 1 million agents internally and must observe and govern all of them.

"The physics of cybersecurity have been completely changed by AI. We build defenses and controls. It was for humans. But humans and agents do not behave the same way and they don't create the same levels of risk," Gallo said.

The other side: Jonathan Black, Former Deputy National Security Advisor of the United Kingdom, offered a counterweight, arguing that existential AI risks represent the 1% while the 99% — deployment into the real economy — is where most risk actually lies. He emphasized that solutions often lie in the technologies themselves and that governments need to partner with both large firms and startups.

Both agreed that AI models don't stop at country boundaries and that governance frameworks must be international. Gallo argued against the false choice between innovation and safety, urging organizations to "move fast and with urgency" on defensive AI adoption.

---

Labor and Faith Leaders Build an Unlikely Alliance

Why it matters: A union leader, a Catholic priest, and a tech policy advocate found common ground that stretches from Bernie Sanders to Steve Bannon — suggesting worker protection from AI could become rare consensus territory.

Where things stand: Lauren McFerrin, Executive Director of the AFL-CIO Technology Institute, described how AI is already dehumanizing workers through algorithmic management, intrusive surveillance, and de-skilling. She highlighted Unite Here's collective bargaining agreement as a concrete model.

"Unite Here, which represents, among many other workers, casino workers in the Las Vegas area, has negotiated provisions in their collective bargaining agreements that gives them up to 6 months notice before any new technology is implemented in their workplace," McFerrin said.

She also flagged the disproportionate impact on women: "AI is coming for a lot of workers, but it's coming for women workers first, actually, and particularly women with high school degrees that do administrative work."

Father Sam Sawyer, Editor-in-Chief of America Media, grounded the discussion in Pope Leo XIV's encyclical Magnificat Humanitas, arguing that work is not merely an economic activity but "a deeply human activity" through which people "participate in God's creative activity."

Wes McEnaney, Labor Outreach and Coalition Builder at the Future of Life Institute, described a pro-human declaration assembly that featured both Bernie Sanders and Steve Bannon, demonstrating that concern about AI's impact on workers transcends partisan divides. He framed the current moment as the culmination of 50 years of deindustrialization and called for more unions and collective bargaining as solutions.

---

Open-Source AI: A Sovereignty Question

Why it matters: As intangible assets grow three times faster than physical ones, the countries and communities that control creative and scientific output through AI models gain a decisive economic and cultural edge.

Where things stand: Dr. Ziad Pourath, Vice President of Imagination and Strategy at the Mozilla Foundation, introduced the concept of the "imagination economy" — the idea that as computing power becomes commodified, the ability to generate and capture new ideas becomes a nation's most important strategic asset. He cited South Korea's cultural exports, now the country's fourth-largest export category at $15 billion, as evidence.

Pourath warned that frontier AI model consolidation threatens to flatten global creativity, noting that 200 million Portuguese speakers in Brazil are barely represented in large frontier models. His framing was blunt: "Open source is a precondition for sovereignty. Already we are in UNGA Week and I've started to hear movements from the frontier labs towards a dilution of that, towards agency and towards leverage as new terms of use. We should wholeheartedly reject any reduction or negotiation around sovereignty."

He added: "It's not just an economic concentration problem. It is an imagination concentration problem as well."

Dr. Maisha Alam, Senior Vice President for Research at FP Analytics, noted that intangible assets have grown threefold faster than physical assets over the past two decades. The discussion covered Rwanda's 2050 visioning exercise, the need for pluralistic sovereign AI ecosystems, and the importance of building distribution pipelines and rights registries to capture value from creative output.

---

Singapore: Build AI Safety Like Aviation Safety

The basics: Singapore has pursued a national AI strategy since 2018, but the post-COVID generative AI boom forced a pivot toward safety infrastructure.

Where things stand: Josephine Teo, Minister for Digital Development and Information of Singapore, described her country's North Star as "AI for the public good, for Singapore and the world" — deliberately prioritizing citizen well-being over commercial adoption speed. She drew an extended analogy to aviation safety.

"If you ask ourselves, how do we get from people not being willing to step on a plane to people being generally unafraid to step on a plane — there is a whole bunch of things that we need to do. It's a very multilayered approach to protection," Teo said.

Singapore has created a testing framework and software toolkit for AI safety, established R&D funding for responsible AI, and convened an international exchange resulting in the "Singapore Consensus." On US-China tech competition, Teo said Singapore avoids dependence on any single ecosystem: "We cannot simply expect to make progress by accessing any single platform or ecosystem. This is not the approach that is sustainable and it doesn't confer resilience. Instead, we try to stay connected with diverse sources of technology and innovation."

What's next: Singapore is about to take over the ASEAN chairmanship and will continue work on the digital economy framework agreement, cross-border data flows, and SME AI adoption.

---

The $50 Smartphone Is Dead

Why it matters: Without affordable devices, AI's benefits remain inaccessible to billions — and the hardware economics are moving in the wrong direction.

Where things stand: Vivek Badrinath, Director General of GSMA, reported that while 96% of the world has mobile coverage (92% in Africa), 3.1 billion people remain offline due to the "usage gap." The primary barrier is handset affordability: a global memory crisis has pushed the cheapest viable smartphone from $50 to $100 — one and a half months' wages for the bottom 20% in sub-Saharan Africa.

Vilas Starr, President and Trustee of the Patrick J. McGovern Foundation, challenged the assumption that connectivity alone solves inequality. "The first is the idea that access equals agency. And that is a foundationally problematic starting point because we have spent billions of dollars to build infrastructure and unfortunately have also created captive user bases for private sector technologies that don't lead to the equitable outcomes we're looking for," Starr said.

Marc-Alexandre Dumba, Minister of Digital Economy and Innovation of Gabon, described his country's mobile-first, super-app approach to e-government, including a national digitalization law requiring procurement preference for local companies and building digital identity onto existing mobile money wallets.

Badrinath discussed the SEALION project (Southeast Asian languages, supported by Singapore and Google/Gates Foundation) and ATLAS-Umoja (African Union) as promising local language AI model initiatives, noting that Google translating an LLM answer is "completely different" from asking a question of a Swahili-native LLM.

---

• TikTok's AI labeling at scale: Tom Varghese, AI Lead for Global Public Policy at TikTok, revealed the platform has labeled over 3 billion pieces of content as AI-generated using watermarking and technical standards. TikTok's AI Literacy Fund, partnering with 20+ organizations including Common Sense Media and the News Literacy Project, produced videos viewed by more than 200 million people in six months. The company also developed "managed topics" — a user-controlled dial for adjusting AI-generated content visibility.
• Finland pitches nuclear-powered AI: Alina Valtonen, Minister of Foreign Affairs of Finland, argued Europe doesn't need to win the frontier AI race but must avoid dependence. She cited Finland's competitive advantages — lowest electricity prices, complete energy independence, Nokia and Ericsson, and a leading quantum computing ecosystem — and noted Google's €13 billion AI infrastructure investment in Finland. Valtonen identified the bottleneck as entrepreneurial risk-taking spirit, not capital or talent, noting most European pension fund money is invested in US stocks.
• Zoom pushes back on data localization: Josh Kalmar, Chief Global Affairs Officer at Zoom Communications, argued digital sovereignty should mean empowerment and autonomy, not geographic data mandates, warning that AI compute demands amplify every unit cost of localization and ultimately limit smaller countries' access to competitive products.
• Philanthropy as risk capital: Melanie Hui, CEO of Luminate, announced a global strategic litigation fund to challenge big tech and described witnessing "algorithmic violence" during the Rohingya genocide. Michelle Jawando, CEO of Omidyar Network, highlighted existing consumer protection frameworks already being applied in California and New Mexico, and New York's first-of-its-kind regulation protecting fashion models from unauthorized digital likeness use.
• FIFA World Cup as tech testbed: Panelists described the 2026 FIFA World Cup as the most-watched event in human history — 6 billion viewers — with Ladan Fakouri, National Security Lead at Lenovo, detailing a Miami command center with dashboards across all 16 stadiums, AI analytics for 48 teams, and 1,248 player avatars. Kate Dillon, Assistant Secretary of State for Cultural and Educational Affairs, highlighted the White House FIFA task force and visa processing improvements.