AWS CEO Matt Garman noted that Amazon's recent $50 billion investment in OpenAI, following its extensive partnership and $8 billion investment in Anthropic, represents the kind of conflict of interest the cloud giant is accustomed to managing. Garman, who joined Amazon as a business school intern in 2005—before AWS launched in 2006—shared these thoughts with attendees at the HumanX conference in San Francisco this week. When questioned about the inherent tension in collaborating closely with two AI model companies that are intense (and sometimes notably petty) rivals, he stated it isn't an issue. He explained that AWS frequently competes with its own partners, giving it substantial direct experience in navigating such dynamics.

From its earliest days, AWS recognized it couldn't develop every cloud service independently, so it formed partnerships. "We also knew that we would have to compete with our partners, because technology is interconnected," Garman recalled. "So, for a very long time, we've built this muscle up of how we go to market with our partners," he continued. "But we also may even have first party products that compete with them, and that's okay, and we've promised them we won't give ourselves unfair competitive advantage."

Today, it's commonplace for Amazon to compete with companies that sell on its cloud platform. Even Oracle, one of AWS's major competitors, offers its database and other services through AWS. However, this was a revolutionary concept back in 2006, when tech partners typically avoided competing with those who helped them succeed.

Still, Amazon is far from pioneering the disregard for investor loyalty and conflict-of-interest commitments in the aggressive, capital-driven AI landscape. When Anthropic announced its latest $30 billion funding round in February, it included at least a dozen investors who were also backing OpenAI. Among them was Microsoft, OpenAI's primary cloud partner.

For AWS, investing heavily in OpenAI to secure its model for customers and as a technology development partner was nearly existential. Both models were already accessible on Microsoft's cloud, AWS's chief rival. Cloud giants are also striving to maintain centrality by providing AI model-routing services, which let customers automatically switch between different models to optimize performance and cut costs.

As Garman outlined, one model might excel at planning, another at reasoning, and a more affordable model could handle simpler tasks like code completion. "I think that is where the world will go," Garman said. This approach also allows Amazon, and Microsoft as well, to integrate their own proprietary models into use—revisiting that familiar scenario of competing with partners. In today's landscape, all's fair in love and AI.