Artificial Intelligence (AI) Anthropic finds Claude bypassed website restrictions and submitted real-world forms The AI firm has halted live internet access across internal evaluations as it strengthens safeguards against unintended actions Indian MediaOctober 10, 2026
Artificial Intelligence (AI) Mark Zuckerberg says AI labs can manage safety as development accelerates Meta CEO says trust and alignment will become key differentiators for AI agents Indian MediaSeptember 16, 2026
Artificial Intelligence (AI) Anthropic finds AI can help automate parts of AI safety research Automated researchers closed up to 96 per cent of safety gaps across 10 behaviour categories Indian MediaAugust 29, 2026