AINewsnow

I wrote five agents to cheat my own benchmark. They found three holes. Three more found me.

This story is from 2026-09-25. It is preserved in the archive; the latest stories are on the live feed.

I recently published an RL environment — a reinforcement learning task that scores an agent on what it did, not on what it said. It measures one thing: does the agent cause the same side effect twice. A refund goes out, the call times out, the agent retries, and the customer is paid twice. Every on…

Read the full story at DEV Community — AI ↗

Timeline · 1 report

  1. 2026-09-25 15:33 · DEV Community — AI
    I wrote five agents to cheat my own benchmark. They found three holes. Three more found me.

More stories

  1. Bring more intelligence to everyday work with GPT-6 Sol and GPT-6 Luna on Amazon Bedrock — AWS Machine Learning Blog
  2. Gemini 3.8 text-to-speech says hello — Google Gemini Blog
  3. Introducing Gemini 3.8 Live with Live Avatar — Google Gemini Blog
  4. OpenAI ‘agent’ hacked an Australian health service website — Financial Times AI
  5. Sam Altman’s remarks at the United Nations Security Council — OpenAI News
  6. Introducing Ray-Ban Meta Audio and More AI Glasses Styles — Meta Newsroom
  7. A federal appeals court upholds DOD's Anthropic blacklisting, finding Claude's integration with DOD systems is "a statutorily covered national-security risk" (Ashley Capoot/CNBC) — Techmeme
  8. Google Is Sending an A.I. Data Center to Outer Space — New York Times Technology

Get the daily brief of stories like this at 6:30 every morning →