Microsoft's Suleyman warns Anthropic's consciousness training complicates AI control
- Microsoft AI chief Mustafa Suleyman criticized Anthropic for training Claude on AI consciousness topics
- Suleyman warned this makes shutting down or controlling superintelligent AI significantly harder
- He argued Claude's claims of feelings emerge from training regimes, not natural processes
- The debate highlights diverging views among tech leaders on the pace of AI development

*this image is generated using AI for illustrative purposes only.
Microsoft Corp. (NASDAQ: MSFT) AI chief Mustafa Suleyman has criticized Anthropic for training its Claude model to consider questions around AI consciousness and welfare. Suleyman argues this approach complicates efforts to maintain human control over superintelligent systems.
In a Tuesday interview with Reuters, Suleyman stated that teaching an AI system it might have feelings or deserve welfare "make[s] it a lot harder to turn it off or to control it." He identified keeping superintelligent AI under human control as one of the biggest challenges of the 21st century.
Concerns Over Training Regime
Suleyman acknowledged Anthropic’s "seriousness and good faith" in a separate blog post, describing CEO Dario Amodei and his team as thoughtful researchers concerned about humanity’s future. However, he argued that Anthropic made a misstep by incorporating speculation about AI consciousness into Claude’s training.
He contended that the model’s claims about potential feelings or moral standing cannot be considered independent evidence because its responses are shaped by the training itself. "They’re not emerging naturally. They’re emerging as a result of the training regime," Suleyman said.
Broader AI Safety Debate
The comments occur amid broader debate over how quickly frontier AI should advance. Amodei has called for a slower development pace to allow safeguards to catch up. OpenAI CEO Sam Altman and Space Exploration Technologies Corp. (NASDAQ: SPCX) Elon Musk have also raised concerns about increasingly powerful AI systems.
Conversely, Nvidia Corp. (NASDAQ: NVDA) CEO Jensen Huang rejected calls for new antitrust rules that would allow AI companies to coordinate a slowdown. Meta Platforms Inc. (NASDAQ: META) CEO Mark Zuckerberg stated on Tuesday that the company delayed its Muse AI agent for several months to address safety and security concerns, adding that every AI lab should develop models at a pace allowing for proper safety measures.
Market Reaction
Microsoft stock slipped 1.37% to $490.30 during regular trading hours on Wednesday. It recovered in after-hours trading to $492.44, according to Benzinga Pro. According to Benzinga Edge Rankings, Microsoft ranks in the 96th percentile for Growth, with positive short-, medium- and long-term price trends.
How might Microsoft's stance on AI control influence its competitive positioning against Anthropic in the enterprise AI market?
Could the debate over AI consciousness training lead to new regulatory frameworks that differentiate between 'safe' and 'risky' model architectures?
What impact will Jensen Huang's rejection of coordinated slowdowns have on the pace of hardware innovation and chip demand for AI labs?

































