OpenAI just proved you can have the best AI in the world and still lose the SEO war to a women's clothing retailer.
The Summary
- OpenAI launched Dots at DevDay, always-on AI agents that work across projects after you close the chat, represented as colorful blob avatars
- The name collides with at least two Chinese AI companies: RedNote's Dots Studio (vision language models) and Shanghai Shengdongshizhang Technology, which filed a US trademark for DOTS in April 2025 covering AI chatbot software
- OpenAI also unveiled a $500 monthly tier to support the premium agent features, positioning Dots to compete with Meta's Muse
- A small fashion company already owns dots.com and its Instagram was quickly invaded by confused users looking for OpenAI's product
The Signal
OpenAI announced over 20 products at DevDay, but Dots represents the company's clearest bet yet on the agent economy. These aren't chatbots that sit idle between conversations. They're designed to proactively tackle work on your behalf, maintaining context and continuing tasks even when you're offline. Sam Altman called it a "premium product" during his presentation, which explains the aggressive $500 monthly pricing tier that comes with it.
The technology matters. The naming disaster matters more right now. When users tried to learn about Dots after the announcement, they hit a wall of confusion that reveals how crowded the AI naming space has become.
"OpenAI's new product name is extremely difficult to look up."
The collision list is remarkable:
- RedNote's Dots Studio launched in 2025, building vision language models
- Shanghai Shengdongshizhang Technology filed a US trademark for DOTS in April 2025, specifically for AI chatbot and personal assistant software
- A women's fashion retailer owns dots.com and the Instagram handle, now fielding questions about AI agents instead of clothing orders
Business Insider notes that internet users are calling out the timing, especially after SpaceXAI recently purchased dot.com. The suggestion: OpenAI either didn't do basic trademark research or decided the name was worth the legal and branding headaches. Neither option inspires confidence in a company trying to position itself as the infrastructure layer for enterprise AI.
This isn't just a branding hiccup. In the agent economy, discoverability is distribution. If developers and enterprise buyers can't easily find documentation, case studies, or community discussions about your product because the name returns irrelevant results, you've kneecapped your own launch. Google still drives how people discover new tools. SEO still matters, even for cutting-edge AI products marketed to technical audiences.
The Implication
Watch how OpenAI handles this. They could rebrand quickly, which would be embarrassing but smart. They could lawyer their way through trademark disputes, which would be expensive and slow. Or they could try to brute-force their way to the top of search results through marketing spend, betting that "OpenAI Dots" becomes distinct enough from "Dots" in six months.
For companies building in the agent space, the lesson is simpler: product names are infrastructure now. Before you announce, buy the domains, check the trademarks in every major market, and search your own product name like a confused customer would. If a fashion retailer or a Chinese AI lab owns the top results, pick a different name. The best agent in the world is useless if nobody can find it.
Sources
Business Insider Tech | Bloomberg Tech | Mashable Tech | Fortune Tech