OpenAI says its board can hold back the release of an AI model even if OpenAI's leadership says it's safe, and announces a new internal safety advisory group
The study of frontier AI risks has fallen far short of what is possible and where we need to be. Ina Fried / Axios : OpenAI touts ‘scientific approach’ to measure catastrophic risk Matthias Bastian / The Decoder : How OpenAI aims to prevent catastrophic AI risks Kwan Wei Kevin Tan / Business Insider : OpenAI is giving its board veto powers over Sam Altman James Farrell / SiliconANGLE : New OpenAI safety team will have power to block high-risk developments Savannah Fortis / Cointelegraph : OpenAI launches ‘Preparedness team’ for AI safety, gives board final say Aamir Sheikh / Cryptopolitan : OpenAI Authorizes Safety Team Against High-Risk AI Developments Amir Efrati / The Information : OpenAI Board Will Review New AI Models, Can Block Releases Maxwell Zeff / Gizmodo : OpenAI's New Board Members Are Now the Boss of Sam Altman (If They Want to Be) Devin Coldewey / TechCrunch : OpenAI buffs safety team and gives board veto power on risky AI Gerrit De Vynck / Washington Post : OpenAI lays out plan for dealing with dangers of AI Abubakar Idris / The Messenger : OpenAI Will Let Board Overturn CEO Sam Altman's Decisions Erin Woo / The Information : After OpenAI Head of Trust & Safety Dave Willner's exit in July 2023, OpenAI has broken up the team, whose vetting procedures some developers said took too long X: @openai : In our safety baselines, only models with a post-mitigation score of medium or below can be deployed, and only models with a post-mitigation score of high or below can be developed further. We also increase security protections commensurate with model risk. [image] @openai : Our Preparedness Team will drive technical work, pushing the limits of our cutting edge models to run evaluations and closely monitor risks, including during training runs. Results will be synthesized in scorecards that track model risk. [image] @openai : Our new safety baselines and governance process will turn these technical findings into safety decisions for model development and deployment. This involves establishing a cross-functional Safety Advisory Group to make safety recommendations. [image] @openai : We are systemizing our safety thinking with our Preparedness Framework, a living document (currently in beta) which details the technical and operational investments we are adopting to guide the safety of our frontier model development. https://openai.com/... @openai : We encourage you to read the full Preparedness Framework (Beta) here for more detail. This is a living document, and we expect to regularly update it as we learn more. https://cdn.openai.com/... LinkedIn: Jonas Schuett : OpenAI just published a framework which describes their process to track, evaluate, forecast, and protect against catastrophic risks from frontier AI models. … Sandro Gianella : 💡 Sharing three interesting projects and initiatives on AI Safety from the OpenAI teams below. …