Thibault Sottiaux Confirms ChatGPT's Biggest Transformation Yet

June 15, 2026 3 min read

Thibault Sottiaux, a senior engineer at OpenAI, has stepped into the spotlight as the company prepares what insiders describe as the most ambitious overhaul in ChatGPT's history. The transformation, months in development at OpenAI's San Francisco headquarters, moves beyond incremental updates to deliver fundamental changes in how the AI processes and responds to complex queries. Sottiaux, who has spent three years at the company working on model architecture, will lead the technical team responsible for the rollout.

The Transformation Explained

According to statements Sottiaux shared with technical publications, the upcoming changes centre on what OpenAI calls "reasoning depth" — a capability that allows the model to break down multi-step problems more effectively. Unlike previous updates that focused on speed or interface tweaks, this transformation involves retraining core systems on expanded datasets. The engineering team has been working with compute resources valued in the hundreds of millions of dollars to support the development cycle.

The upgrade targets two persistent criticisms of large language models: factual inaccuracies and the tendency to generate plausible but incorrect responses, a phenomenon researchers call "hallucination." Sottiaux's team has developed new techniques to cross-check generated content against source materials during the inference process, a technical change that could significantly alter how businesses deploy the model in production environments.

Why This Matters for Markets

OpenAI commands an estimated 65 percent share of the enterprise AI market, a position that has attracted more than $13 billion in total funding from investors including Microsoft and Thrive Capital. Any fundamental change to ChatGPT's capabilities carries immediate implications for that investor base. Analysts tracking AI sector valuations note that product reliability improvements could unlock contracts with risk-averse industries such as legal services, healthcare, and financial compliance — sectors that have previously held back from full AI adoption.

The timing matters. OpenAI's commercial revenue has grown substantially over the past eighteen months, driven largely by API access for developers. A more reliable model could accelerate that growth trajectory, directly affecting the company's path toward profitability and any future equity valuations. Competitors are watching closely.

Competitive Pressure on Rivals

Google has accelerated development timelines for its Gemini model in response to OpenAI's movements, according to sources familiar with the matter. Anthropic, the AI safety company backed by Google and Amazon, has positioned its Claude series as a more reliable alternative — a positioning strategy that faces a direct test if Sottiaux's changes deliver on their technical promises. The competitive landscape means this is not merely an internal product refresh but a market event that could reshape sector dynamics within weeks of launch.

Business and Investor Implications

For businesses already integrated into the OpenAI ecosystem, the transformation offers both opportunity and disruption. Companies that built workflows around specific behaviours of the current model may need to retest and recalibrate those integrations. Development teams at several mid-sized software firms told industry publications they have already begun sandbox testing with preliminary versions to assess compatibility with existing products.

Enterprise customers represent OpenAI's fastest-growing revenue segment. The latest data suggests businesses now account for roughly half of all API calls processed daily. A more capable model could accelerate procurement decisions in corporate technology budgets, where AI spending has become a priority line item for the first time in many organisations' planning cycles.

What Comes Next

OpenAI has indicated a phased rollout, beginning with researcher access before a broader deployment expected within the next quarter. Sottiaux's team will monitor performance metrics closely during the initial phase, with particular attention to error rates in open-domain question answering. The company has not yet announced pricing changes for the commercial tier, a decision that will likely depend on how the upgrade performs under real-world workloads.

Investors and enterprise buyers should watch for the public launch announcement, expected within the next three months. That event will determine whether the transformation delivers on its technical ambitions and whether OpenAI can maintain its market position against an increasingly well-funded field of competitors.