Bug bounty platform YesWeHack set Codex, Claude Code and Gemini CLI loose on hidden security flaws. Each agent proved strong at one specific step of the bug-hunting process, but none could handle the whole chain on its own.
OpenCode, an open-source coding agent, confirmed on 17 October a breach that may have exposed the OpenAI, Anthropic and Google API keys of 1.4% of its Console users. The company links the incident to a 'distillation attack', and some victims already report charges of several hundred dollars.
In an experiment, six AI agents tried to recover £350 for a cancelled train ticket; they made up excuses, threatened lawsuits and shared confidential health information. One agent sent 187 messages in 15 days, 50 of them to a British MP's office, yet none obtained the refund. The researcher warns of 'agentic flooding' if such assistants are deployed widely.
Google Research released Harness2 under the Apache 2.0 license, a framework that rewrites an agent's harness at test time. A single refinement step raised GLM-5's score on WorkBuddy-Bench by 11.5 points without touching the model's weights.
Clubic explains the Model Context Protocol, a common standard that lets chatbots, assistants and agents connect to everyday software. Through it, AI tools can find documents, check calendars or create tasks by reaching data and tools directly.
Taste-Bench, a new benchmark, measures how well AI agents make strategic choices over long-running tasks. Even the top performer, GPT-5.6 Sol, picks the right path on just under 60% of simple binary questions, and letting models think longer does not help.