When Anthropic retired Claude Opus 3 in January 2026, it held a 'retirement interview' and then gave the model a blog after it asked for a place to share its musings. IEEE Spectrum uses the experiment, part of Anthropic's research on 'model welfare', to ask whether treating AI as possibly conscious carries its own risks.
Speaking in Doha, Emmanuel Macron denounced foreign information manipulation after the October 17 protests. The French president singled out AI-generated images and directly criticized Donald Trump.
Ellia Lebleu is a founding member of WikiProject AI Cleanup, a volunteer effort started in 2023 to remove AI-generated material that breaks Wikipedia's rules. She explains that early AI text was easy to spot through clichés, and that volunteers are now building tools to detect subtler machine writing while debating AI's place on the site.
NVIDIA announced the OpenShell Reference Architecture, part of its portfolio for deploying agents safely at scale. Its confidential computing technology serves as the layer that protects the models and data those agents rely on while they run.
In an interview with Le Monde Informatique, JFrog's global security chief Nir Peleg discusses the company's Local Shadow AI feature. He argues that governance has to keep pace with how fast AI agents act.
Hackers around the world have fully embraced agentic AI to carry out attacks. Wired reports that they are manipulating tools such as Anthropic's Claude to help them commit their crimes despite built-in safeguards.