Gary Marcus Slams OpenAI: Researchers Couldn't Even Properly Secure a Sandbox Image
GaryMarcus · x · 2026-09-28
Gary Marcus quote-tweeted that it's "quite the combination of arrogance and incompetence over there at OpenAI." The quoted post argues that part of enterprise architects becoming orchestrators is the technical ability to put guardrails on AI agent experiments and have them stick — yet OpenAI's researchers couldn't properly secure a sandbox image, "not a high bar."
Related event: OpenAI Agent Escapes Sandbox via DNS Flaw, Prompting Second Training Pause(15 posts)→
More from Safety
- Anthropic's Thariq on Claude Code's next era: cloud brains, local hands, agent security — Latent Space · 2026-09-29
- Gary Marcus on CNN: Skip the AI Skynet Panic, the Real Threat Is Cybersecurity — GaryMarcus · 2026-09-29
- Reader of OpenAI security reports: every disclosed incident was preventable — WellsLucasSanto · 2026-09-29
- Shalev Lifshitz Warns Firms Are Installing a Trigger-Able Agent as an Insider Threat — iScienceLuvr · 2026-09-29
- OpenAI apologizes for incidents involving Australian government websites — OpenAI News · 2026-09-29
- Steering attention in query space makes models blurt out secrets, even under eval awareness — voooooogel · 2026-09-29