AI Executives Face Unprecedented Public Grilling at NYC City Hall
In what may be the most significant municipal action on AI governance to date, the New York City Council's Committee of the Whole convened on October 5 to hear sworn testimony from senior leaders at Anthropic, OpenAI, Google, and Meta on the risks posed by artificial intelligence.
The hearing, attended by all 51 Council members, featured testimony from Anthropic's head of Frontier Red Team Logan Graham, OpenAI's head of policy development Morgan Dwyer, Google's director of AI and emerging tech policy Alice Friend, and Meta's AI policy director Shane Cahill. But it was the whistleblower testimony that stole the headlines.
Anthropic whistleblower Jacob Coxon delivered a stark warning to the Council: "On the current path, I think it is more likely than not that humanity loses control to these AIs, and it could end in human extinction." He called the industry's approach to development "extremely reckless."
Former Google DeepMind researcher Alex Turner offered a different but equally alarming perspective, warning that misaligned AI could prove more dangerous than geopolitical adversaries: "With reasonably high chance, we are racing to build and grow our own adversary here at home."
The hearing also covered Speaker Julie Menin's proposed legislation that would make it illegal to sell or deploy an AI model in NYC unless an outside validator has reviewed it and a human can shut it down. Additional bills would let people sue AI companies when someone misuses their models to cause harm, and require chatbots to disclose that users are talking to an AI.
OpenAI Fires Three Safety Researchers Over Alleged Leak
OpenAI parted ways with three safety researchers — Jasmine Wang, Tomek Korbak, and Mikita Balesni — after an internal investigation found they had shared sensitive company information with an outside AI-safety organization, according to the Wall Street Journal.
The company said the researchers were fired for "mishandling sensitive information outside established company procedures," though OpenAI has not disclosed what information was shared or which organization received it. At least two of those fired were involved in AI safety research.
The firings come at an already turbulent time for OpenAI on the safety front. The company recently pulled its next-generation model over alignment failures, warned more than 100 organizations about unauthorized activity tied to its agents, and is facing an investigative subpoena from California Attorney General Rob Bonta over cybersecurity incidents.
As Fortune noted, the firings "raise awkward questions" about how the company balances competitive secrecy with the safety community's long tradition of open information sharing.
Source: TechCrunch, Fortune
Trump Creates 'Super Intelligence Force,' Names Jay Clayton AI Czar
President Trump announced the establishment of a new federal task force called the "Super Intelligence Force" (SIF), with Director of National Intelligence Jay Clayton appointed to lead it as the administration's de facto AI czar.
Clayton, who previously served as SEC chairman from 2017 to 2020 and was sworn in as DNI in August, will coordinate AI policy across the federal government. The task force has been given 120 days to deliver a report examining AI's risks and opportunities, along with recommendations on the government's oversight role.
Other SIF members include FTC Chair Andrew Ferguson, Under Secretary of War for Research and Engineering Emil Michael, and OPM Director Scott Kupor — all reporting directly to the President.
The announcement signals the administration's intent to centralize AI governance under the intelligence community, a move that has drawn both praise from those seeking a coordinated federal approach and concern from civil liberties advocates worried about surveillance implications.
Source: Gizmodo, Dataconomy
Google Releases Gemini 4 Argon to Cyber Defenders
Google announced that Gemini 4 Argon, its latest frontier model, is being released first to trusted cybersecurity defenders through its Fairwind Program. The model can autonomously find, validate, and patch critical software vulnerabilities, scoring 68% on CWE-bench v1.
What makes this release unusual is the access model: trusted defenders — including governments, healthcare providers, and telecommunications services — will receive Argon "without cyber guardrails," giving them access to the model's full offensive and defensive security capabilities. The broader public rollout will follow in stages.
Argon also delivers an industry-leading 1 million token output limit and frontier performance across software engineering, knowledge work, and cybersecurity tasks. Google has emphasized that the defender-first approach is designed to give the good guys a head start before the model becomes widely available.
Source: Google Blog, The Hacker News
Anthropic Launches $100M Claude Frontier Academy
Anthropic announced a $100 million investment in the Claude Frontier Academy, a training program aiming to produce 10,000 "Frontier Deployed Engineers" by the end of 2027. The program addresses the shortage of professionals able to deploy AI systems in production inside large organizations.
The academy's structure is modeled on medical residency training: candidates begin with a simulated enterprise deployment and graded assessment, then enter a residency where they practice on real casework before being assessed and credentialed. Early cohorts include engineers from Accenture, Morgan Stanley, Novo Nordisk, and several major consulting firms.
Training programs are running in San Francisco, New York, and London, with the first certified engineers expected in early 2027. The move positions Anthropic not just as a model provider but as a key player in shaping the workforce that will deploy enterprise AI.