Report questions trust in OpenAI CEO as company pitches AI policies

In short: OpenAI released a set of public policy ideas about how AI should be managed, while a major report raised new questions about whether CEO Sam Altman can be trusted to follow through.

What's going on

OpenAI published recommendations it says could help society if AI systems become extremely capable. The company said it wants rules that “keep people first,” and it pointed to risks like AI being used to undermine democracy or getting out of human control.

The same day, The New Yorker published a long investigation based on interviews with more than 100 people familiar with Altman’s work, plus internal memos and multiple interviews with Altman. The article said some current and former insiders see a pattern where Altman tells people what they want to hear, then changes course later. The New Yorker reported no single definitive piece of proof, but it cited messages from former leaders who said the pattern made them worry about safety and governance.

OpenAI’s policy ideas include trying shorter workweeks, creating a “public wealth fund” (a shared pool of money that could pay citizens a dividend), and taxing automated work to help fund programs like Social Security and Medicaid. It also suggested tighter controls for a small set of the most powerful AI models, especially those that could increase risks in areas like cyberattacks or dangerous chemicals.

What to watch

Public trust is becoming a central issue for AI companies, especially as governments and businesses use their systems more. Watch whether lawmakers respond with stricter AI safety rules, and whether OpenAI changes its leadership or oversight to address concerns about who is steering the company.

Source: Arstechnica