Sam Altman admits OpenAI ‘was on the wrong side of history in open source debate’

Sam Altman (19459055), CEO of OpenAI made a On Friday,made a striking admission that his company had been “on the wrongside of history” in regards to open source AI. This could signal a seismic shift in strategy with China’s competition intensifying and open models gaining traction.

This candid admission came during a Reddit discussion. Just days after Chinese AI company DeepSeek launched its “Ask Me Anything ” session OpenAI’s R1 model, which is open source and claims to be comparable in performance with OpenAI systems at a fraction the cost, has rattled global markets.

Yes, we are talking about [releasing model weights] Altman wrote. “I personally believe we have been on a wrong side of history and need to figure out another open source strategy.” he said, noting that not everyone at OpenAI shared his view and that it wasn’t the company’s current top priority.

This statement is a significant departure from previous statements. Openai’s increasingly proprietary approach in recent years has been criticized by some AI researchers and former partners, most notably. Elon Muskis suing the company, claiming that it has betrayed its original mission of open source.

Sam Altman on DeepSeek : ‘We’ll maintain less of a leading’

Altman’s comments come amid DeepSeek’s appearance caused market turmoil. The Chinese company’s claim of building advanced AI models at just Training costs of $5.6 million Nvidia stock was sent to a new high The market value of the U.S. companies plummeting was the largest drop in a single day in US history.

Altman admitted in the same AMA that “we will produce better models but we will maintain a smaller lead than in previous years,” addressing DeepSeek directly.

Sam Altman admits that OpenAI’s closed-door strategy may be flawed.

DeepSeek’s breakthrough, regardless of whether its specific claims are accurate, has highlighted changing dynamics in AI development. The company claims to have achieved its results by using only 2,000 Nvidia GPUs– far fewer than the estimated 10,000+ chip typically deployed by major AI laboratories.

The approach suggests that algorithmic innovations and architectural optimization may be more important than raw computing power. This revelation threatens OpenAI’s entire business model, which is built on exclusive access massive computational resources.

The open source debate – innovation vs. safety

But DeepSeek has also grown. National security concernsThe company stores data on servers located in mainland China where government access could be possible. NASA is the latest U.S. agency to restrict the use of the application, citing “security concerns and privacy concerns.” The company was established in 2015 as a non profit with the mission to ensure artificial general intelligence benefits mankind. Open source advocates have criticized its transition to a capped-profit model and increasingly closed strategy.

The correct reading is that “Open source models are exceeding proprietary models,” wrote Meta’s chief AI scientists Yann Lecun, ten LinkedInresponding to DeepSeek’s emergence. “They built new ideas on top of the work of others. Everyone can benefit from their work because it is open source and published. “Open research and open source” is the power behind open source.

A new chapter in AI

Altman’s comments indicate that a strategic change may be on the horizon, but he stressed that OpenAI is not prioritizing open source at this time. This hesitation reflects a complex reality that AI leaders face: balancing security, innovation, and commercialization within an increasingly multipolar AI environment.

OpenAI’s bottom-line is not the only thing at stake. The decision of the company could have a profound impact on the entire AI ecosystem. Open-sourcing key AI models could accelerate innovation, democratize access and improve AI safety and security. However, it could also complicate efforts for OpenAI to achieve its mission of ensuring AI safety and cybersecurity.

Altman’s admission came after DeepSeek’s shock to the market, not before. This suggests that OpenAI is reacting rather than leading market forces. This is a dramatic role reversal from a company who has long been positioned as AI’s North Star.

As DeepSeek’s debut settles, one thing becomes apparent: the real disruption doesn’t just concern technology or market value. It’s about challenging assumptions that closely guarded AI model is the surest way to artificial general intelligence. Altman’s admission could be less about being wronged by history than it is about acknowledging that history has changed.