Anthropic CEO Dario Amodei criticized rivals OpenAI and Google for their reactive "code red" emergencies amid the AI race, highlighting unnecessary panic. Instead, Anthropic focuses on enterprise applications for steady progress and safety. He warns against reckless overspending, advocating a prudent strategy for sustainable AI development.
A 2025 AI safety report reveals top models like Google's Gemini, Anthropic's Claude, and OpenAI's ChatGPT barely pass evaluations with C averages, exposing vulnerabilities to misuse, bias, and manipulation. Amid fierce competition and emerging risks, experts urge stronger safeguards to ensure reliable integration into critical sectors.
|