OpenAI merges ChatGPT and Codex under Greg Brockman. The side quests are over.

OpenAI co-founder and president Greg Brockman has permanently taken charge of the company’s product strategy, merging ChatGPT, Codex, and the developer API into a single product organisation. In an internal memo seen by Wired, Brockman wrote that OpenAI will “invest in a single agentic platform and to merge ChatGPT and Codex into one unified agentic experience for all.” The restructuring formalises an interim arrangement that began in early April when Fidji Simo, OpenAI’s CEO of AGI deployment, went on medical leave.

The reorganisation places four pillars under Brockman. Thibault Sottiaux, the engineer who built Codex into one of OpenAI’s fastest-growing products, now leads core product and platform across consumer, enterprise, and developer surfaces. Nick Turley, who oversaw ChatGPT’s expansion to more than 900 million weekly active users, moves to a role focused on enterprise products and critical industries. Brockman retains his existing responsibility for AI infrastructure, including the Stargate data centre programme. OpenAI told TechCrunch that Simo collaborated with Brockman on these changes and is expected to return, though no timeline has been disclosed.

The consolidation is the culmination of a strategic retreat that began in December when CEO Sam Altman declared a “code red” and told staff the company needed to refocus on its core ChatGPT experience. Since then, OpenAI has shut down Sora, its video generation app, which consumed vast computing resources relative to its revenue and triggered the collapse of a planned $1 billion Disney investment. It shelved its adult mode for ChatGPT after pushback from staff, advisors, and investors. It halted OpenAI for Science. Kevin Weil, head of Sora, and Bill Peebles, who led the product, both departed. The company has described these paused initiatives internally as “side quests.”

Brockman himself explained the compute constraint driving the consolidation on a recent podcast, saying that OpenAI’s computing power is “not enough for even a personal assistant and the Codex line.” When you are resource-constrained, you cannot maintain separate product teams, separate roadmaps, and separate engineering organisations for products that are converging toward the same capability. The merger eliminates that redundancy and concentrates engineering effort on a single surface that can handle conversation, code generation, tool use, and autonomous task execution.

The competitive context is immediate. Cursor has reached $2 billion in annualised revenue and is in talks to raise at a $50 billion valuation, demonstrating that agentic coding is the fastest-growing category in developer tools. Anthropic’s Claude Code has been gaining ground with enterprise developers. Google’s Gemini has grown its share of AI web traffic from 5.7% to 21.5% over the past twelve months, according to SimilarWeb, while ChatGPT’s share declined from 86.7% to 64.5%. Google I/O 2026 opens on Monday with agentic coding and Gemini updates headlining the agenda. Two years ago, OpenAI countered Google I/O by unveiling GPT-4o the day before the conference. This year, OpenAI is not countering with a product launch. It is countering with an org chart.

The restructuring is also IPO-facing. OpenAI is preparing to go public in Q4 2026, targeting a valuation of approximately $852 billion. A simpler product story, one platform rather than a portfolio of separate apps, is easier to present to institutional investors. It also creates a cleaner revenue narrative: one subscription tier, one developer platform, one enterprise offering, all building on the same underlying model infrastructure. The previous structure, in which ChatGPT pursued consumer reach while Codex served developers and the API monetised the ecosystem separately, created internal competition for compute, engineering talent, and strategic attention.

The timing is complicated by the Musk v. Altman trial, which began jury selection on Monday in Oakland federal court. Musk’s lawsuit seeks up to $150 billion in damages and the unwinding of OpenAI’s nonprofit-to-profit conversion. The most damaging evidence is a 2017 diary entry from Brockman himself, who wrote: “I cannot believe that we committed to non-profit if three months later we’re doing b-corp then it was a lie.” Brockman is now simultaneously leading OpenAI’s product strategy, overseeing its infrastructure buildout, and serving as a central figure in the trial that could determine whether the company’s legal structure survives.

The unified platform, which OpenAI is internally describing as an agentic “super app,” will roll out gradually. Codex will expand first to cover productivity tasks beyond coding before ChatGPT and the company’s research tool Atlas are folded in. No launch date has been announced. The ambition is a single application where a user can have a conversation, write code, execute multi-step tasks, browse the web, manage files, and interact with external services, all within one interface, all powered by the same model, and all generating revenue through a single subscription or API billing relationship.

Whether one person can run both product strategy and infrastructure for a company with 900 million weekly users, thousands of employees, a multi-billion-dollar data centre programme, and an IPO on the horizon is a question OpenAI’s board has implicitly answered by giving Brockman both jobs. The answer is that there is nobody else they trust to do it. The side quests are over. What remains is the main quest: build the platform, ship the agent, go public. Everything else has been cut.