In today’s AI race, breakthroughs are no longer measured in years—or even months—but in weeks. But Z.ai’s lead didn’t last long, either. Just as Anthropic had been undercut by GLM-5’s release, GLM-5 ...

Microsoft researchers discovered that poisoned AI models exhibit normal behavior until specific trigger words cause them to ...

Anthropic says it is testing a powerful new AI model that can spot serious weaknesses in software, and releasing it as part ...

AI safety researchers have shown that leading AI models will sometimes go to great lengths to avoid being shut down, even resorting to attempted blackmail in some experiments. Now it turns out these ...