Google Had ChatGPT-Like AI a Year Before OpenAI but Refused to Launch It — and the Decision Reshaped AI History

In a revelation that has ricocheted across the tech industry, a former Google DeepMind engineer who now leads core products at OpenAI has publicly confirmed that Google had built an internal chatbot functionally equivalent to ChatGPT nearly a full year before OpenAI launched ChatGPT in November 2022. The claim brings a fresh insider perspective to one of the defining what-if questions of the AI era: why did the company that arguably invented the technology decide not to ship it?

What an OpenAI Insider Confirmed

Thibault "Tibo" Sottiaux, a former Google DeepMind engineer who now leads Codex and core products at OpenAI, stated plainly on X (formerly Twitter) that he was part of the team that developed Google's internal conversational AI chatbot. The project was initially called LMChat and later renamed under another codename .

Sottiaux described it as "basically ChatGPT one year before it came out" . His post responded to AI engineer Cheng Lou, who had recalled an earlier interview in which Google Chief Scientist Jeff Dean mentioned the company had an internal chatbot but felt it was no better than traditional Google Search .

The Reason Google Held Back: Fear of Cannibalizing Search

Sottiaux offered a blunt explanation for why the product never saw the light of day: "Google was too nervous to release it, and DeepMind was blocked from shipping products that could disrupt Google's business" .

This aligns with multiple earlier reports and even Google CEO Sundar Pichai's own admissions. Pichai has acknowledged that Google had built ChatGPT-like technology early but cited safety and quality concerns for delaying its release . Internal discussions also centered on the fear that a conversational AI chatbot would cannibalize Google's core search advertising business, which generates the vast majority of the company's revenue .

Sottiaux later added in another exchange that he thinks "a lot" about Google's decision not to release the chatbot .

The Competitive Impact: A Classic Innovator's Dilemma

Google's strategic caution had consequences that reshaped the competitive landscape of AI:

• OpenAI seized first-mover advantage. By launching ChatGPT in November 2022 while Google held back, OpenAI defined the modern conversational AI category, attracted hundreds of millions of users, and set the product and UX standards that every competitor has since had to follow .

• Google was forced into a reactive posture. The company rushed Bard (later rebranded Gemini) to market months after ChatGPT's debut. The launch was widely seen as a hurried response rather than a deliberate product rollout, and it suffered early credibility issues .

• It reshaped industry dynamics. The episode is now frequently cited as a textbook case of the "innovator's dilemma" — in which a dominant incumbent with superior technology declines to disrupt its own revenue model, allowing a smaller, hungrier competitor to capture the market. Google's hesitation cost it the lead in conversational AI despite having the underlying technology first .

Former Google engineer Blake Lemoine, who was part of Google's AI team until 2022, has also said that Google was on the verge of releasing a chatbot in the fall of 2022 — roughly the same time as ChatGPT — but executives delayed it over safety concerns .

A Moment That Defined an Era

Sottiaux's confirmation, combined with earlier admissions from Google leadership and other former employees, paints a clear picture: Google had the technology to dominate conversational AI from the start. The decision to protect its search business at the expense of a potentially disruptive new product allowed OpenAI to not only catch up but define the category. The question now is whether Google's belated push with Gemini can undo the damage from that missed window.