Editorial illustration for OpenAI's DevDay Push: Locking Developers Into Its Platform, Not Just Models
OpenAI's DevDay: Locking Developers Into Platform
OpenAI brought its DevDay roadshow to Berlin this week, and most of what got announced had already been said in San Francisco. The one genuinely new piece of news: Dots, OpenAI's AI assistant, should become more widely available across Europe "in a few weeks," according to Romain Huet, the company's head of developer experience. That's pending talks with regulators, and for now Personal Pro accounts across the European Economic Area, Switzerland and the UK still can't get in, though some business accounts already have access.
The other numbers OpenAI did share point to where its priorities sit. Weekly active Codex users in Germany have grown twentyfold since January 2026, the company said, building on a sevenfold jump reported back in June. No absolute figures were given, as usual.
Coding tools and agents are clearly getting the resources, while video input models and Sora have been sidelined over cost and demand concerns. Underneath the recycled headlines, the real story in Berlin was about infrastructure, developer lock-in, and OpenAI's push to make itself the default platform rather than just the model underneath it.
Beyond the individual product updates, the event showed how tightly OpenAI wants to lock developers and businesses into its ecosystem. With the Agents API, OpenAI runs the agent infrastructure itself.
Why this matters
Berlin DevDay was framed as a feature rollout, but the real announcement was architectural. Dots arriving in Europe and Codex usage jumping twentyfold in Germany since January matter less as metrics than as evidence of OpenAI's actual strategy: get developers building on its rails, not just calling its models. That's a different kind of lock-in than the usual API dependency we've warned about.
Once your coding agents, your assistant integrations, and your internal tooling all assume OpenAI's platform conventions, switching costs stop being about model quality and start being about rewriting your stack. The Sora shutdown and the missing video-input model tell the same story from the other direction, OpenAI will cut unprofitable product lines fast, which should worry anyone building long-term plans on top of features that don't pay their own compute bills. For founders and engineering leads, the question isn't whether GPT-5 or Codex is good enough today.
It's whether you're comfortable building on infrastructure whose roadmap is set by OpenAI's margins, not your dependency on it.
Related Reading
Browse all OpenAI news →
AI Tools & Apps Claude gains shared context in Excel, PowerPoint; Microsoft adds Copilot Cowork
AI Tools & Apps Sarvam Signs 5-Year Defence AI Deal Covering Army, Navy and Air Force
AI Tools & Apps LG's recent webOS update adds Microsoft Copilot app, now removable
Related Topic OpenAI researcher quits, citing distrust over ad‑driven engagement metrics
LLMs & Generative AI
Related Topic OpenAI launches GPT-Image 1.5 with precise editing for enterprise visuals
LLMs & Generative AI
From Archives Macmillan Workers Say Publisher Celebrated an AI Assistant Before Killing It
From Archives Judge Sees AI in Courts as Tool for Greater Access to Justice
From Archives OpenAI Reportedly Plans to Purchase Google AI Chips
Open Source
From Archives Oracle and OpenAI Connect ChatGPT Plugins to Enterprise Data and Cloud
LLMs & Generative AI
Common Questions Answered
What is OpenAI's strategy with the Agents API according to the article?
OpenAI's Agents API represents a shift toward platform lock-in by having OpenAI run the agent infrastructure itself rather than allowing developers to manage it independently. This architectural approach aims to lock developers and businesses into OpenAI's ecosystem beyond just using its models for API calls.
When will OpenAI's Dots AI assistant become available in Europe?
According to Romain Huet, OpenAI's head of developer experience, Dots should become more widely available across Europe in a few weeks, pending regulatory approval. Currently, Personal Pro accounts across the European Economic Area, Switzerland, and the UK still cannot access Dots, though some business accounts have limited access.
How does OpenAI's lock-in strategy differ from traditional API dependency?
Rather than just creating dependency through API calls to its models, OpenAI is building a comprehensive platform where developers' coding agents, assistant integrations, and internal tooling all run on OpenAI's infrastructure. This represents a deeper architectural lock-in that makes it harder for developers to switch platforms once their entire workflow is built on OpenAI's rails.
What evidence does the article cite of OpenAI's platform expansion in Germany?
The article mentions that Codex usage has jumped twentyfold in Germany since January, demonstrating OpenAI's growing presence in the European developer market. This metric serves as evidence of OpenAI's strategy to get developers building on its platform infrastructure rather than just using its models.
Further Reading
- DevDay 2026 Recap - OpenAI
- OpenAI DevDay 2026: Dots, Sign in With ChatGPT and Lock-In - The NovTech
- OpenAI Introduces Always-On Dots Agents - Reworked
- OpenAI DevDay 2026: Key Announcements for Teams - ScrumLaunch
