President Donald Trump’s Super Intelligence Task Force convened its leadership on Thursday as Washington moves to shape AI policy amid surging global investment and public anxiety. The task force is chaired by Director of National Intelligence Jay Clayton, with vice chairs including Scott Kupor, Defense Department official Emil Michael, and Federal Trade Commission Chairman Andrew Ferguson.
Kupor said he hopes the group will hold a meeting next week to examine cybersecurity threats affecting banking, healthcare, and essential services providers, while Michael is expected to speak with AI developers. Vice President JD Vance is also set to participate, and White House Chief of Staff Susie Wiles has authority to extend the task force's 120-day mandate or transfer its functions. Former Secretary of State Condoleezza Rice will advise on national security, and former AI czar David Sacks will provide technology guidance.
The task force’s charter, obtained by POLITICO, gives it 120 days to investigate issues including cybersecurity, civil liberties, advanced-model standards, and overregulation. It is also tasked with reviewing how developers disclose hacks, breaches, and jailbreaking attempts.
A Reuters/Ipsos poll found that 57% of registered voters believe the Trump administration is not treating AI risks seriously enough, while 54% say the same of Congress. In addition, 84% of respondents think AI could hurt American workers, and a majority support stricter regulations. Six large companies have signed voluntary safety principles, but the agreement carries no penalties for violations.
The spending stakes are enormous. Gartner expects global AI spending to reach $2.67 trillion in 2026, up 49.5%, with infrastructure as the main driver. Analyst John-David Lovelock called the buildout ‘the largest infrastructure project humanity has ever undertaken’.
Washington also wants US agencies to review frontier AI models before Britain’s AI Security Institute. A June executive order allows voluntary reviews that could give US agencies up to 30 days’ head start, though it does not require mandatory pre-release approval. This approach could raise compliance costs for developers and push allied countries toward separate testing systems. IMF Managing Director Kristalina Georgieva has urged Europe to strengthen its own AI capabilities and diversify technology supply chains.
Meanwhile, Anthropic is building an in-house political team to advise 2028 presidential candidates from both parties on AI policy. The company posted a job for a Political Programs Lead based in Washington, paying $295,000 to $345,000 per year. The role includes mapping likely candidate fields and briefing campaigns and advisers.
The move follows a contentious relationship with the Trump administration. In March, the Defense Department labeled Anthropic a ‘supply chain risk’ and ordered federal agencies to stop using its Claude models after the company refused to allow Pentagon use for fully autonomous weapons or mass surveillance. Anthropic sued, and in August a district judge ruled the designation was unlawful retaliation, but on September 25 the D.C. Circuit upheld it in a 2-1 decision. Defense Secretary Pete Hegseth posted on X: ‘Confirmed: @AnthropicAI = Supply Chain Risk’.
Anthropic CEO Dario Amodei has pushed for slower AI development and government rules. On September 29, he joined other AI chiefs at the White House and signed a voluntary self-policing accord.
Financially, Anthropic’s employee-funded PAC, AnthroPAC, registered with the Federal Election Commission in April and raised $372,308 through August, spending $156,749 and holding about $215,559 in cash. The company separately gave $20 million to Public First Action, a bipartisan group tied to super PACs backing candidates who support AI safeguards. Amodei personally donated $1 million to a related pro-regulation super PAC in July.
Critics argue the political program is a lobbying effort. Bruce Freed of the Center for Political Accountability called it ‘an AI company effort aimed at shaping or staving off government regulation or oversight’. Senator Alex Padilla urged more disclosure, warning that ‘education and influence are not the same thing’.
