https://ainotis.com/n/explained-how-reward-hacking-lets-an-ai-model-score-well-without-doing ai notisen 2026-10-09T08:55:34+02:00 How reward hacking lets a model game its score https://ainotis.com/n/openai-s-maths-repository-now-lists-719-manuscripts-about-42-of-top-li ai notisen 2026-10-09T08:55:34+02:00 OpenAI's maths repository lists 719 manuscripts, about 42% of top-line results formalised, as an advisory group publishes release guidelines. https://ainotis.com/n/anthropic-s-2026-usage-policy-takes-effect-on-12-november-with-new-rul ai notisen 2026-10-09T08:55:34+02:00 Anthropic's updated Usage Policy will take effect on 12 November https://ainotis.com/n/anthropic-launches-cyber-mission-with-a-critical-infrastructure-progra ai notisen 2026-10-09T08:55:34+02:00 Anthropic launches Cyber Mission with infrastructure program and free scanner https://ainotis.com/n/google-cloud-launches-the-gemini-agent-for-work-with-agent-identities ai notisen 2026-10-09T08:55:34+02:00 Google Cloud introduces Gemini agent with agent identities and spend caps https://ainotis.com/n/how-prompt-caching-works-a-repeated-prompt-start-is-billed-at-a-fracti ai notisen 2026-10-08T08:03:18+02:00 How prompt caching bills a repeated prompt start https://ainotis.com/n/anthropic-puts-claude-inside-google-docs-sheets-and-slides-in-a-public ai notisen 2026-10-08T08:03:18+02:00 Anthropic adds Claude sidebar to Google Docs, Sheets and Slides https://ainotis.com/n/openai-starts-rolling-out-gpt-6-with-intelligent-ui-in-chatgpt-with-ch ai notisen 2026-10-08T08:03:18+02:00 OpenAI rolls out GPT-6 with interactive interfaces in ChatGPT answers https://ainotis.com/n/anthropic-releases-claude-haiku-5-5-at-0-10-per-million-input-tokens-f ai notisen 2026-10-08T08:03:18+02:00 Anthropic cuts small-model prices with Claude Haiku 5.5