Your AI vendor is now a brand statement
Last week, ChatGPT uninstalls surged 295%. In a single week. The trigger wasn't a bug, a price hike, or a bad update. It was a deal: OpenAI's agreement to let the US Department of Defense use its AI in classified settings.
Users didn't wait for clarification. They just left.
What happened
OpenAI CEO Sam Altman admitted the negotiations were "definitely rushed." Anthropic, which had been publicly reprimanded by the Pentagon for refusing to loosen its safety restrictions, got excluded from the deal entirely. While OpenAI was shaking hands with the DoD, Anthropic quietly launched memory upgrades for Claude and released a tool specifically for importing data from rival chatbots, aimed squarely at AI switchers.
Claude's downloads went up as ChatGPT's fell.
This used to be a tech story. It's a brand story now.
The AI tool you use is a visible choice
Three years ago, nobody asked which cloud provider you used. Nobody cared which email service processed your company's mail. Infrastructure was invisible.
AI is different. The AI tools a company uses are becoming visible to clients, employees, and the public. When a firm puts "Powered by ChatGPT" on their product, that's a statement now. Some customers will read it as efficiency. Others will read it as a choice to work with a company that signed a military AI deal in a rush.
This isn't theoretical. The 295% uninstall surge came from regular consumers, not activists. Enterprise decisions move slower than consumer ones, but the same values calculus is starting to show up in procurement conversations. The people sitting across from you at client meetings deleted the app over the weekend.
In London, hundreds of protesters marched through King's Cross (home to OpenAI, Meta, and Google DeepMind's UK offices) chanting "Pull the plug! Stop the slop!" It was the largest anti-AI protest yet in Europe. These aren't just fringe signals. They're a measure of how far the conversation has moved beyond product features.
Three risks worth thinking about
Client-facing AI exposure
If you use an AI tool in client deliverables (reports, summaries, research, drafts), the vendor behind that tool is part of your service now. A client who deleted ChatGPT from their personal phone may ask pointed questions when they see it powering your workflow. Know which tools you're using and be ready to explain why.
Employee alignment
The anti-AI protests weren't just a consumer story. A real segment of knowledge workers cares deeply about which AI companies their employer legitimizes. If your internal stack becomes public (and internal stacks have a way of becoming public), it can affect hiring and retention, especially among younger talent.
Vendor lock-in and vendor risk
The switching cost for AI tools is higher than people realize. You've trained your team on one tool, built prompts around its quirks, maybe integrated it into your CRM. Anthropic now offers a purpose-built import tool to make switching from ChatGPT easier, which tells you something about how real this churn is. If you've built deep dependencies on a single AI vendor, you now carry a new kind of exposure: reputational risk tied to your vendor's decisions.
What to do with this
This isn't a call to abandon any particular tool. Most AI vendors will have controversial moments. OpenAI has had plenty before this one.
The point is to treat AI vendor selection the way you treat other business decisions that carry public exposure.
Know your stack. Document which AI tools touch client work versus internal-only work. Keep them separate.
Build portability into your workflows. In practice, this means keeping your prompts in plain text files your team owns rather than locked inside a vendor's interface, and avoiding integrations that can't export your data. If your outputs live entirely inside one vendor's ecosystem, you have no leverage when that vendor does something that conflicts with your values or your clients' expectations.
Develop a position. At some point a client or employee will ask which AI tools you use and why. Probably sooner than you expect. Having a clear answer ready is table stakes for professional services now.
The AI vendors are fighting their own brand wars. Whether your business ends up as collateral damage in someone else's controversy depends on how much thinking you've already done about where you stand.
The 295% figure will fade from headlines. The underlying dynamic won't.