Microsoft AI CEO Mustafa Suleyman Says Anthropic Is Worsening Safety Risks

Microsoft AI CEO Mustafa Suleyman recently claimed that Anthropic is actively exacerbating industry safety threats, proving that tech giants are getting desperate....

Feed
September 18, 2026
Microsoft AI CEO Mustafa Suleyman Says Anthropic Is Worsening Safety Risks


Tech giants love lecturing us about responsibility. They preach caution while sprinting to ship whatever half-baked model they can monetize first. Yet the narrative took a bizarrely aggressive turn recently when Microsoft AI CEO Mustafa Suleyman decided to point a finger directly at a rival lab. According to Suleyman, AI safety threats are very real, and Anthropic is making the whole situation significantly worse by moving too fast or perhaps just not playing by Redmond's unwritten rules.

It is a fascinating piece of corporate theater. Microsoft, a company deeply entangled in a multi-billion-dollar marriage with OpenAI, is suddenly worried about someone else pushing the envelope on safety. The irony is staggering. When you are sitting on top of an empire fueled by rapid deployment, complaining about a competitor's pace feels less like a genuine ethical awakening and more like a tactical maneuver to control the regulatory narrative. Everyone wants guardrails, naturally, provided those guardrails happen to block their rivals more than themselves.

Microsoft AI CEO Mustafa Suleyman Says Anthropic Is Worsening Safety Risks

We need to talk about what is actually happening beneath the PR spin. Building complex software has always required discipline, constraint, and a healthy respect for unintended consequences. Instead, the current AI field resembles a high-stakes poker game where the players keep changing the rules of the table while demanding medals for being cautious. Suleyman might be right that the underlying risks demand our attention. But pointing the blame outward while your own ecosystem scales at breakneck speed misses the forest for the trees.

Real engineering requires sweat and accountability, not finger-pointing in tech policy debates. Small teams building actual products don't have the luxury of spending millions on lobbying or public spats over who cares more about safety. Just, we have to live with the systems we deploy. If the people steering the trillion-dollar ships spent half as much time writing solid code as they do manufacturing rivalry, maybe the rest of us wouldn't have to navigate this constant storm of hype.