Which AI vendor is going to come out on top: Anthropic, Google, Microsoft, OpenAI?
At the moment, they’re all great at general tasks, and the gap between them keeps shrinking. So the real competition has moved somewhere less glamorous: into your calendar, your inbox, your Salesforce instance, and your legal review queue.
In other words, the battle is about who becomes indispensable to how your organisation actually works.
The Spotify lesson
Spotify has shown us how it could pan out. In 2008, every music service had the same catalogue. The same songs were available everywhere, at the same price. Today, Spotify dominates. The platform enabled users to create and share playlists. They were yours. It meant the product got better every month you didn’t cancel. It became harder to cancel because you would lose years of context that quietly shapes every session.
AI platforms are doing a similar thing
Microsoft, Google, and others are running basically the same playbook. By embedding AI into tools such as Power Platform, Teams, and Workspace, they’re embedding themselves into your workflows and your institutional memory. The longer you stay, the more they know about how your business runs.
If you replace customer service agents with an AI system, the knowledge lives in configurations, integrations, and months of fine-tuning. Ripping it out would become an infrastructure project rather than a simple software licensing decision.
What this means for your AI strategy
If you’re choosing a vendor, you need to consider which platform will know our business best in 18 months, and what will it cost us to leave?
If you’re automating a workflow, document it. If you ever want to migrate, you need to know exactly what you’ve built and how it works.
The next dominant AI company will probably win because it holds the workflows nobody wants to rebuild. That is a less exciting story than AGI, but it is also the one that is actually happening.
Leave a Reply