YouCongress

Comment by Joe O'Brien

Researcher on the Frontier Security team at the Institute for AI Policy and Strategy (IAPS), focused on AI assurance, reliability, crisis preparedness, and developer transparency.
Models used internally at AI companies have capabilities beyond those of publicly-available models, so it's important that risks from these models are reported externally. [...] We argue that whenever a substantially more capable or riskier model is deployed internally, the developer should create a risk report and argue why the model is safe to deploy.
Unverified (Apr 29, 2026)
Source:
Joe O'Brien, “Internal AI Model Risk Reporting Framework for Companies,” LinkedIn post, April 29, 2026. The post’s LinkedIn activity timestamp corresponds to April 29, 2026; O'Brien writes that higher-capability or riskier models deployed internally should have a risk report and a safety case.
Like Share on X 31min ago

Policy proposals and claims

AI labs should conduct risk assessments before high-stakes internal deployment of frontier models
votes For
Statement relation verification history Unverified Report this
No statement relation verification comments yet.
Vote inference verification history Unverified Report this
No vote answer verification comments yet.
replying to Joe O'Brien

Other authors to follow

OpenAI AI research organization Dario Amodei CEO at Anthropic Anthropic AI safety research company Yoshua Bengio AI Pioneer, Turing Award winner Sam Altman CEO at OpenAI Demis Hassabis Nobel laureate, AI Researcher and CEO of DeepMind António Guterres UN Secretary-General Bernie Sanders U.S. Senator from Vermont Pope Leo XIV Head of the Catholic Church