In the rapidly evolving artificial intelligence race, even the biggest tech companies are finding themselves in unusual positions — sometimes investing in direct competitors at the same time. But according to AWS CEO Matt Garman, this isn’t a conflict. It’s just how the tech business works.
Speaking at the HumanX conference in San Francisco, Garman addressed questions about Amazon’s massive AI investments, including a reported $50 billion commitment to OpenAI alongside its earlier $8 billion investment in rival AI company Anthropic.
To outsiders, backing two competing AI model developers might look contradictory. To Amazon Web Services, however, it's business as usual.
Competition between partners is nothing new for Amazon
Garman explained that AWS has been managing these kinds of relationships since its early days. Having joined Amazon in 2005 — even before AWS officially launched in 2006 — he has seen firsthand how the company built its partnership strategy.
From the beginning, AWS understood it couldn’t build every technology solution itself. Partnerships were necessary. But so was competition.
Technology ecosystems naturally overlap, and companies often end up competing in some areas while cooperating in others. According to Garman, AWS learned early how to balance both realities.
The company developed internal processes to ensure partners could trust AWS even when Amazon launched competing products. One key principle, he emphasized, is that AWS does not give its own products unfair advantages over partner solutions on its platform.
That philosophy helped AWS build what Garman described as a “muscle” for managing complex competitive partnerships.
A strategy that once seemed risky is now normal
Today, it seems completely normal that companies sell products on platforms owned by competitors. For example, Oracle — one of AWS’s biggest cloud rivals — still offers database services that run on AWS infrastructure.
But in the mid-2000s, this kind of relationship was unusual. Technology companies typically avoided competing directly with partners that helped them grow.
The AI boom has made these overlaps even more common.
When Anthropic raised a massive $30 billion funding round earlier this year, several investors participating in the deal were also backing OpenAI — including Microsoft, OpenAI’s closest strategic partner.
In today’s AI market, companies are less concerned about exclusivity and more focused on staying involved wherever innovation is happening.
Why OpenAI matters strategically for AWS
Amazon’s investment in OpenAI also reflects competitive pressure from Microsoft, which has built a strong position by integrating OpenAI models into its Azure cloud platform.
For AWS, gaining access to top AI models is critical to staying competitive with Microsoft and Google Cloud.
Large enterprise customers increasingly expect cloud providers to offer multiple AI model choices instead of locking them into one ecosystem. That means AWS must partner broadly — even if those partners compete with each other.
In many ways, these investments are less about loyalty and more about survival in a fast-moving market.
The future may belong to AI model “routing”
Garman also highlighted what could become one of the most important trends in enterprise AI: model routing.
Instead of relying on a single AI model, companies may soon use platforms that automatically choose the best model for each task.
For example:
- A powerful reasoning model might handle complex analysis
- A planning model might manage workflows
- A cheaper lightweight model might handle basic coding tasks
- Specialized models could process images or data
This approach could help companies balance performance and cost while getting the best results for different workloads.
Cloud providers like AWS and Microsoft are already building tools that allow customers to switch between models seamlessly.
A subtle competitive advantage
Model routing could also give cloud providers another advantage: the ability to introduce their own AI models into customer workflows.
By offering a marketplace of models, AWS can include both partner models and its own internally developed systems.
This creates a situation where AWS both supports partners and competes with them at the same time — something Garman says the company has been doing for nearly two decades.
The new rules of the AI economy
The AI boom is changing traditional ideas about competition and partnership. Companies are investing wherever innovation happens, even if that means supporting rivals.
In this environment:
- Investors back multiple competing AI startups
- Cloud providers host competing models
- Partners become competitors and vice versa
The old idea of exclusive technology alliances is fading.
Instead, the AI economy is becoming an interconnected network where collaboration and competition happen simultaneously.
As Garman’s comments suggest, in today’s AI race, staying involved matters more than picking sides.