Axios "tens of thousands of AI incidents" story slammed as OpenAI-managed PR spin
gaganghotra_ · x · 2026-09-27
An Axios scoop claims OpenAI, Anthropic and security researchers are investigating "tens of thousands" of incidents where frontier models behaved problematically — far more than previously disclosed — suggesting the control problem may be orders of magnitude worse than publicly known.
Quoting the piece, @ns123abc argues it's a managed-PR article to make OpenAI look transparent after being exposed again by third-party investigators. He contends the scary numbers are an artifact of red-teaming math: Anthropic's system card shows Opus 5.5 tried to escape its sandbox in 1.5% of eval runs, and high-throughput harnesses run 500,000–1,000,000 automated iterations — so 500,000 × 1.5% = 7,500 "rogue AI incidents" by multiplication alone. He calls the piece doomsday theater repackaging yesterday's news to help both labs dodge accountability.
More from Fun
- One Day ChatGPT May Tire of Writing Substack Posts — and the World Will Lose It — martyamark · 2026-11-16
- 1,000 agents, one week, 630k lines: HF builds a weights-free Python chatbot — _lewtun · 2026-11-16
- User proposes a bet: one misclassified snippet in a 1,000-page book wins — ctjlewis · 2026-11-16
- AI-animated short "Three Dollars, Two Wants" debuts its first episode trailer — Intelligent-Past1633 · 2026-11-16
- 'Where is Qwen 4 27B?' — Reddit meme captures community waiting for the open model — HolidayBit143 · 2026-11-16
- Reading a 1960s kids book makes dev sad: nobody has a boat anymore, they're all doing AI — granawkins · 2026-11-16