Summary:
Original Link:
Original Article:
Here are the most important things that happened in AI risk management this week.
🔗 Learn more: https://lnkd.in/ez-SQbkw
Policy news:
▶︎ EU rules on GPAI models started to apply: https://lnkd.in/eUP-gVg5
▶︎ European Commission and AI Board formally adopted the GPAI Code of Practice: https://lnkd.in/etWT5PMY
Company updates:
▶︎ OpenAI released GPT-5: https://lnkd.in/e_h7MNdG
▶︎ OpenAI also released gpt-oss, an open-weight reasoning model: https://lnkd.in/erUkUCWA
▶︎ Anthropic released Claude Opus 4.1: https://lnkd.in/eSkmCtzb
▶︎ Anthropic published framework on developing safe and trustworthy agents: https://lnkd.in/eWnEduSH
Research:
▶︎ FMF report on third-party assessments: https://lnkd.in/eFpehrGQ
▶︎ METR evaluation results of GPT-5: https://lnkd.in/eEt_g76M
▶︎ Position paper on openness and AI safety: https://lnkd.in/e3eq2Ch5
▶︎ GovAI paper on regulating downstream AI developers: https://lnkd.in/eSvZuh_C
Job opportunities:
▶︎ European Commission is looking for experts to join the Scientific Panel on AI: https://lnkd.in/exiBf8Sh
▶︎ SaferAI is looking for a Policy Associate: https://lnkd.in/epQ9Pmeh
▶︎ GovAI is looking for a (Senior) Research Editor: https://lnkd.in/eNyvQQGM
▶︎ GovAI is looking for a Hiring Manager and Hiring Specialist: https://lnkd.in/eaA9GRYF
👉 Subscribe to the AI Risk Management Newsletter for weekly updates: https://lnkd.in/ezEwqVwn