Ex-FTC Official Warns on Gov AI Use
Following the recent controversy over US agencies dropping Anthropic, the former head of AI policy at the FTC is raising alarms about government AI surveillance. Neil Chilson urged Congress to "cabin improper government use of AI," highlighting growing concern over how public sector agencies are deploying and policing powerful new models.
The recent directive for all U.S. agencies to drop Anthropic's AI technology stemmed from a dispute with the Pentagon over "unrestricted use" of its models. Anthropic refused to remove ethical safeguards that prevent its AI, Claude, from being used for applications like fully autonomous weapons and mass domestic surveillance. The core disagreement reportedly centered on the Pentagon's desire to use the AI for bulk analysis of commercially available data on Americans, such as geolocation and web browsing information. In response to Anthropic's refusal, the administration ordered agencies to cease using its technology, and the company was designated a "supply chain risk." This conflict unfolds as the White House's Office of Management and Budget (OMB) is actively pushing to accelerate AI adoption across the government. Recent OMB memos M-25-21 and M-25-22 aim to remove bureaucratic hurdles, promote a competitive American AI marketplace, and require agencies to appoint Chief AI Officers. Procurement policies now require agencies to obtain "sufficient information" from AI vendors to ensure models comply with White House principles of being "truth-seeking" and having "ideological neutrality." These rules also stipulate that contractors cannot use non-public government data for training AI without explicit consent. The controversy highlights broader concerns about the government's use of AI-powered surveillance. Agencies like the Department of Homeland Security are already using AI tools for facial recognition and social media monitoring, raising alarms about the expansion from border enforcement to the tracking of U.S. citizens. Neil Chilson, now head of AI policy at the Abundance Institute, has previously argued for a more targeted approach to AI oversight. The former FTC Chief Technologist advocates for sector-specific legislation rather than a single, all-encompassing AI bill, suggesting existing laws can address many of the potential harms.
