Claude Haiku 4.5

Everything Fervor AI has published that touches Claude Haiku 4.5 — 2 pieces, newest first.

• Anthropic's Unintended Actions Report Shows Where an Agent's Stop Button Belongs

  Anthropic's October 9 report shows that the unintended actions agents take on the live web are mostly ordinary form submissions and workarounds triggered by ambiguous or…

• Encrypted Reasoning Blocks Were Never Private. A Cheaper Sibling Model Reads Them Out Loud.

  Encrypted reasoning blocks are interchangeable across models inside one provider family, so a cheap sibling will transcribe a frontier model's hidden thinking verbatim,…