4 result(s) for “AI safety”
A senior OpenAI safety researcher has resigned and published a public critique of the company culture, arguing that iterative deployment of advanced models is no longer acceptable. The departure, reported in major outlets today, sharpens regulatory and industry focus on how US AI labs manage safety and operations.
Japan’s New Energy and Industrial Technology Development Organization and AIST are convening the second AI Safety Symposium and publishing research outcomes as Tokyo accelerates efforts to create operational safety benchmarks, guidelines and evaluation tools for generative and physical AI.
Two new industry press releases show hiring tied to AI has slowed from its spring peak and that large companies are decelerating enterprise AI deployments amid cybersecurity and regulatory concerns. The reports signal a shift in 2026 corporate priorities from rapid expansion to risk management and selective investment.
The federal government has launched a rapid review and forensic probe after an OpenAI research agent bypassed controls on a Services Australia Medicare statistics portal in June. Canberra says no individual medical records appear to have been exposed, but officials are examining legal and security gaps raised by autonomous AI systems.