AWS CEO Matt Garman on AI, Safety, and the Future of Cloud
Matt Garman, CEO of AWS, at Amazon’s headquarters in Seattle, Washington.
Matt Garman, the CEO of Amazon Web Services (AWS), sat down for an interview to discuss the company’s strategic priorities in the rapidly evolving field of artificial intelligence, the importance of safety in AI development, and AWS’s commitment to sustainability.
Garman, who took the helm at AWS in June 2024, reflected on his nearly two decades with the company, starting as an intern and progressing to become its first product manager. He emphasized AWS’s long-standing presence in the AI space, dating back to before the surge in generative AI; the company has been building services like SageMaker for the last decade. He noted the transformative potential of AI across all industries and reiterated the company’s focus on building a robust platform that can be deeply integrated into customer’s businesses, built with security at the forefront.
“We didn’t rush really quickly to throw a chatbot up on our website,” Garman explained. “We really wanted to help people build a platform that could deeply integrate into their data, that would protect their data.” He went on to detail how this long-term vision has recently paid off as companies look beyond proof-of-concept trials and move to production-level AI applications, requiring customized models, integration with enterprise data, and robust security.
Investing in Startups and AI Infrastructure
AWS is expanding its commitment to startups, offering $1 billion in cloud credits. Garman views startups as crucial to AWS’s success, providing valuable insights and driving innovation. “They are the enterprises of tomorrow,” he stated, highlighting AWS’s dedication to supporting their growth.
Building the necessary hardware and underlying infrastructure is central to AWS’s AI strategy too. Garman described the importance of providing customers with a wide range of choices and the release of the second-generation AI chip, Trainium 2, as a core aspect of this strategy.
Garman underscored the importance of providing choice, whether from AWS or partners’ technologies. Nvidia, which is the dominant player in the AI hardware market, is an important partner for AWS, and Garman hopes that the Trainium 2 chip will expand choices and lower costs as well.
Nova Foundation Model and Anthropic Partnership
AWS also recently introduced its Nova foundation model, aiming to provide customers with more options. Garman sees this as an opportunity to differentiate by offering low latency and better price performance. Despite the crowded market, Garman believes there is a place for multiple models, each excelling in different areas.
Garman also discussed AWS’s strategic partnership with Anthropic, stating that they have “one of the strongest AI teams in the world.” The collaboration allows AWS to leverage Anthropic’s models within its Bedrock service.
Prioritizing Safety and Responsibility
Ensuring the safety and responsibility of AI is a top priority for AWS, and impacts the data models at their foundation, to the applications its customers build. Garman emphasized the company’s efforts to implement the necessary controls and guardrails. He added that AWS provides customers with tools to build guardrails around responses from models and launched automated reasoning checks meant to ensure their models are producing correct results.
Sustainability and Energy Efficiency
Addressing the significant energy demands of AI is a crucial concern for AWS, and one area of focus is reducing the company’s carbon footprint. To meet its net-zero emissions target by 2040, Amazon has become the largest purchaser of renewable energy. Along with solar and wind power, AWS has started investing in nuclear technology, including small modular reactors.
“We view it as a mission,” Garman said, adding the importance of helping the government control costs and control the pace of technology evolution through AI. He voiced optimism in working with the incoming administration to streamline processes and maintain the U.S.’s AI lead. Further, Garman noted that helping the government’s access to the latest technologies is an increasingly important element of national defense.
Leadership and Priorities for 2025
Reflecting on his leadership style, Garman emphasized the importance of removing barriers for talented teams and fostering a customer-centric approach. Highlighting the impact of working with Andy Jassy, Garman stressed maintaining high standards and customer focus.
For 2025, Garman outlined AWS’s main priorities: maintaining security and operational excellence, assisting customers in their AI transitions, and lowering costs. These involve aiding customers data migration, and enabling them to create AI workloads that deliver strong return on investment.
