OpenAI Confronts Anthropic Threat: A Crucial Strategic Shift

⚡

Key Takeaways

1OpenAI abandons Sora, its video application, to focus on its flagship products in light of increased competition from Anthropic.

2The one billion dollar partnership with Disney is canceled, freeing up resources for more strategic projects.

3Anthropic is experiencing rapid growth, narrowing the revenue gap with OpenAI thanks to its enterprise tools and Claude Code.

💡Why it matters — OpenAI needs to realign its strategy to maintain its position against innovative and aggressive competitors.

OpenAI Reorients Its Strategy to Counter Anthropic

Sora, the Video App Set Aside

OpenAI has recently decided to terminate its Sora project, a video generation application, amid accusations that the company was spreading itself too thin. Fidji Simo, head of applications at OpenAI, emphasized that the company was trying to juggle too many projects at once, which was detrimental to its core products. Launched in the fall of 2025, Sora initially generated considerable excitement before its appeal waned.

Jérôme Marin explains that this initial success occurred during a period when OpenAI was adopting an excessively diversified approach. This strategy, driven by a sense of invincibility, ultimately led to mixed results. With only 5% of ChatGPT users opting for a paid subscription, OpenAI must urgently increase its revenue. Of the 900 million claimed weekly users, only a tiny fraction contributes financially. Sora, resource-intensive and lacking immediate monetization prospects, has become difficult to justify.

End of Partnership with Disney: A Strategic Surprise

The abandonment of Sora is not limited to the consumer app. OpenAI has also decided to end its professional segment, which surprised many observers. The company had indeed signed a major partnership with Disney, involving a billion-dollar investment and the use of over 200 characters for video creations. This partnership is now canceled, along with the ambitious project to create a fully AI-generated film, the future of which remains uncertain.

Each abandoned Sora video frees up GPUs for projects deemed more strategic. Jérôme Marin specifies that this decision responds to a dual logic:

Freeing up resources for software development, which currently represents the core of OpenAI's business.
Reallocating talent towards creating models intended for businesses, rather than focusing on video.

Three Warning Signals for OpenAI

OpenAI has faced several warning signals that motivated this strategic repositioning, after being overtaken by the competition.

GPT-5: A Disappointing Launch

The first warning signal came with the launch of GPT-5 in the summer of 2025. The public was not convinced by this new version, which did not deliver the expected performance improvements. Critics pointed to a perceived cold tone, persistent hallucinations, and performance sometimes inferior to that of GPT-4o, forcing OpenAI to urgently reintegrate the older model.

Google’s Gemini: Increased Competition

Two months later, a second signal emerged with Google launching its Gemini 3 model, which was universally praised. This launch confirmed that OpenAI's lead had significantly diminished, prompting Sam Altman to refocus efforts on OpenAI's core activities.

Anthropic: A Rapid Commercial Rise

The most concerning signal came from the rapid growth of Anthropic. Their success, particularly with Claude Code, has been spectacular. Launched in May 2025, this coding tool reached an annualized revenue of one billion dollars in just six months, then 2.5 billion by February 2026.

The revenue gap between OpenAI and Anthropic is closing quickly. Anthropic now reports 19 billion dollars in annualized revenue, while OpenAI claims 25 billion, of which about 70% comes from consumer subscriptions to ChatGPT. OpenAI's strategy of offering ChatGPT to businesses is no longer sufficient. Companies are seeking more concrete tools with a clear return on investment.

OpenAI: From Leader to Follower

OpenAI, long seen as the pioneer of AI, is witnessing a shift in its brand image. ChatGPT was the first AI experience for many users, but this advantage is eroding.

Today, many companies are first turning to Anthropic to explore their solutions. This change in perception is difficult to quantify but has a major strategic impact. OpenAI finds itself in an unprecedented position, that of a follower in the professional segment, even though it remains a leader in the consumer space.

After three and a half years of ChatGPT, the business model justifying massive investments in computing power seems to be shaping around agency and code, areas where Anthropic excels. Jérôme Marin emphasizes that, contrary to expectations, it is not assistants or copilots that have become the main drivers of the AI business, but rather agency and computer code.