New AI benchmarks could help developers reduce bias in AI models, potentially making them fairer and less likely to cause harm. The research, from a team based at Stanford, was posted to the arXiv ...

AI development must focus on creating a scalable, efficient and ethical framework for seamless integration into daily life.

Google has done away with references to 'diversity' and 'equity' on its Responsible AI team page in a nod to the Trump administration.

Model weights form DeepSeek-R1’s operational ... Developers can integrate fairness metrics and allow users to flag problematic outputs. Explainable AI (XAI) methods help reveal how DeepSeek ...