Sam Altman on why OpenAI went from 'open' to 'closed' source code

CEO Sam Altman says OpenAI doesn't publicly release the source code of its AI models to protect their "safety threshold."

During a Reddit Q&A session over the weekend, Altman was asked by a user: "OpenAI has moved from an open source approach to a more closed model in recent years. Can you explain why this changed?"

“I think open source plays an important role in the ecosystem. There are many great open source models in the world today,” the CEO replied. “But closed source with powerful, easy-to-use APIs and services plays an equally important role. That’s what we’re good at, and that way we can more easily achieve the desired security threshold.”

Altman then added that he was proud of the value people were getting from OpenAI 's existing services . "We want to open up more things in the future," he stressed.

Sam Altman, CEO of OpenAI. Photo: TechCrunch

According to Business Insider , Altman made the above comments after OpenAI was criticized for no longer publicly releasing the source code for AI models like GPT-4o and OpenAI o1. Until November 2019, the company still publicly released GPT and GPT-2.

OpenAI started out as a startup focused on developing AI that was non-profit and open, as the name suggests. However, since 2019, the company has partially shifted to for-profit - the reason why one of its founders, Elon Musk, has left.

In December 2022, after ChatGPT launched, Musk wrote on X that calling OpenAI a nonprofit was “no longer correct.” In March, the American billionaire sued Altman and OpenAI, accusing them of breach of contract and abandoning their original mission, and said he would drop the lawsuit if OpenAI changed its name to “ClosedAI.” Musk withdrew the lawsuit in June, then filed a new one in August, claiming he was “tricked” into founding OpenAI.

According to Bloomberg , the choice of open or closed source is a hot topic in the AI ​​industry. One of OpenAI's competitors pushing open source is Meta with its Llama model, the latest being Llama 3.2 released in September. Meta CEO Mark Zuckerberg said he believes open source is "safer than other options" and "necessary for a positive AI future."

Also in the Reddit Q&A, Altman said OpenAI's GPT-5 model "probably won't be released this year." He also admitted to using ChatGPT to answer some questions online.

Bao Lam

Sam Altman: 'AI will be smarter than humans in the next few thousand days' 70
Sam Altman Admits OpenAI Doesn't Understand How GPT Works 15
Elon Musk compares Sam Altman to 'Game of Thrones' character
Sam Altman: 'I Would Choose a Different Name for OpenAI If I Could Go Back in Time'