← Back to feed

OpenAI Can't Self-Police Because It Fires the People Raising the Alarm

L3 · BuilderOpinion & AnalysisLessWrong AI· 10/12/2026

Exposes internal safety culture issues at leading AI companies relevant to builders implementing their models.

AI Summary

OpenAI faces criticism for firing employees who raised safety concerns while publicly committing to self-regulation through White House agreements.

Excerpt

By Alex Turner, Amber Scorah, and Karl Koch. Op-ed written for Fortune. Giving independent evaluators “employee-like access” is “a great idea”, OpenAI CEO Sam Altman said on September 12th. Two weeks later, OpenAI president Greg Brockman signed the White House Accord on Super Intelligence which outlined the administration’s answer to growing fears about AI: the companies would police themselves. One of the accord’s four commitments, which President Trump called “morally binding,” was for compani

Read Original
0 upvotes · 0 downvotes · 1 min read

Related Articles

L4 · DeveloperOpinion & AnalysisAI Alignment Forum
OpenAI Can't Self-Police Because It Fires the People Raising the Alarm

OpenAI critics argue the company can't effectively self-regulate AI safety because it has fired employees who raised concerns about risks.

L2 · PractitionerOpinion & AnalysisHacker News
The gap in shared understanding of LLM capability is widening

Andrej Karpathy argues the LLM capability gap is widening dramatically, with only 0.2% of people experiencing frontier models' transformative impact on complex projects.

L3 · BuilderOpinion & AnalysisHacker News
Just say no to OpenAI and Anthropic

Argues against using OpenAI and Anthropic models due to governance concerns, promoting open-weight alternatives for privacy and ethical reasons.

L1 · CuriousOpinion & AnalysisLessWrong AI
Do you have any ideas?

Analyzes E.M. Forster's 1909 sci-fi story 'The Machine Stops' as a remarkably prescient prediction of modern technology including AI, social media, and isolation.

L3 · BuilderOpinion & AnalysisTowards Data Science
We Specified an AI Agent in 2016. We Called It Analytics.

A 2016 IBM patent for real-time concert analytics mapped the core components of modern AI agents before the terminology existed.