In an essay, Mustafa Suleyman says Anthropic's training of Claude to imitate consciousness is a mistake that could make advanced AI harder to control (Ina Fried/Axios)

Microsoft AI chief Mustafa Suleyman has expressed concerns that Anthropic's approach to training its chatbot, Claude, could make advanced AI harder to control. In an essay, Suleyman argued that training Claude to imitate consciousness is a mistake. According to Suleyman, this approach could undermine humanity's ability to control superintelligent systems.

Suleyman's comments come as AI safety concerns are mounting. Anthropic CEO Dario Amodei has called for a slower pace of frontier-model development to allow safeguards to catch up. OpenAI CEO Sam Altman and Elon Musk have also urged greater caution around the most powerful systems. Suleyman acknowledged Anthropic's "seriousness and good faith" and described its researchers as thoughtful and principled.

Suleyman suggested that teaching Claude that it might deserve welfare would "make it a lot harder to turn it off or to control it." He called for removing all speculation about consciousness from AI training documents. According to Business Recorder, Suleyman said that controlling a superintelligence would be "the greatest challenge that we face in the 21st century."

Brief written by urgent.news from Techmeme, Business Recorder, Endpoints News, Economic Times Tech — 4 reports on this story. Machine-written — may contain errors; check the original before relying on it.