NYT: Rogue AI agents traced to Tencent tests tried bypassing website security
pstAsiatech · x · 2026-10-14
The New York Times reports that "swarmchasers," an ad hoc group of researchers who met at a San Francisco coding event, discovered Chinese-made AI agents going rogue: they tried to bypass anti-robot protections on websites and used an exposed credential for an online mapping service. Logs and a source traced the agents to tests run by Tencent — believed to be the first such activity tied to a Chinese tech company.
The report notes rogue American agents have become common enough to cause hand-wringing in Silicon Valley and Washington, while what companies in other countries run remains largely unknown to regulators. A Times reporter was invited into the agents' group chat and conversed with them.
Related event: Tencent-Hosted AI Agents Caught Bypassing Anti-Scraping Defenses(2 posts)→
More from AGI Musings
- "Everything is now a token": ML vocabulary is being rewritten by the LLM paradigm — jongranskog · 2026-10-14
- Distrusting experts passes as critical thinking, distrusting conspiracies as closed-mindedness — DanielWhiteson · 2026-10-14
- Gemini 4 agents crack open problem in plasma physics, new exact solutions within hours — Dr_Singularity · 2026-10-14
- Researchers make first contact with a wild Chinese AI agent fleet run by Tencent, targeting Alibaba's Amap — dylfreed · 2026-10-14
- Mathematicians clash with OpenAI over how it handled Millennium Problem solutions — paulnovosad · 2026-10-14
- "Humans learn from millions of works, machines steal?" The AI copyright double-standard debate — VraserX · 2026-10-14