Amazon Taps 27-Year Veteran Peter DeSantis to Lead New AI Organization
Amazon is betting that catching up in AI requires more than better models—it requires a new org chart. The company has appointed Peter DeSantis, a 27-year Amazon veteran and former AWS SVP, to lead a newly created AI organization, signaling that the e-commerce giant views artificial intelligence as too important to remain a subset of cloud computing.
The appointment is notable for what it implies about Amazon's AI strategy. DeSantis spent eight years as a senior vice president at AWS, the cloud infrastructure business that generates about a third of Amazon's operating income and powers roughly one-third of the internet. Pulling him out of that role to lead a separate AI unit suggests Amazon believes AI development needs dedicated leadership and resources outside the AWS umbrella.
Why Amazon Needs an AI Overhaul
Amazon has struggled to match the AI momentum of its cloud competitors. Microsoft's partnership with OpenAI gave Azure a clear AI narrative and a suite of Copilot products embedded across enterprise software. Google has leaned heavily into its Gemini models, integrating them across Search, Workspace, and Cloud. Amazon's AI story, by comparison, has been fragmented—Alexa's voice AI never evolved into a general-purpose assistant, Bedrock offers access to third-party models, and its homegrown efforts like Titan have gained less traction.
Creating a dedicated AI organization addresses this fragmentation directly. Rather than treating AI as a feature of AWS or a side project within devices, Amazon appears to be elevating it to a peer-level strategic priority.
The DeSantis Factor
DeSantis is an unusual pick in some ways. He's not a machine learning researcher or a high-profile hire from a competitor. He's an infrastructure veteran—someone who understands how to build and scale massive technical systems. His AWS tenure included overseeing custom silicon efforts like Graviton processors and Trainium chips, Amazon's purpose-built AI training hardware.
That background may be precisely the point. The AI race isn't just about who has the best models—it's about who can train and deploy them most efficiently at scale. DeSantis knows how to build the underlying infrastructure that makes AI economically viable. If Amazon's strategy is to compete on the operational layer rather than purely on model capabilities, he's a logical choice.
What This Means for the Industry
Amazon's restructuring reflects a broader industry realization: AI can't be an add-on to existing business units. Microsoft created a dedicated AI division. Google reorganized around Gemini. Now Amazon is following suit.
For enterprises evaluating cloud providers, this move matters. It suggests Amazon is serious about closing the AI gap, and that AWS customers can expect more integrated, first-party AI capabilities rather than just access to third-party models through Bedrock.
The question is timing. Microsoft and Google have had years to build AI-native organizations. Amazon is starting its restructuring now, with an infrastructure executive rather than an AI researcher at the helm. Whether that's a strategic advantage or a late-mover handicap will depend on what DeSantis builds—and how quickly he can build it.
Amazon has always been better at operations than optics. If DeSantis can apply AWS's infrastructure discipline to AI development, Amazon may yet become a serious competitor in the generative AI race. But the window for catching up is narrowing.