OpenAI Assembles New Safety Team Led by CEO Sam Altman, Starts Training on Next AI Model

OpenAI has announced the formation of a new Safety and Security Committee following the departure of top executives earlier this month.

This move comes as the company begins training its next flagship AI model, designed to succeed the widely used GPT-4 technology.

The new committee is led by CEO Sam Altman and includes several high-profile directors and experts.

OpenAI Reveals New Safety and Security Committee

(ILLUSTRATION) This illustration photo produced in Arlington, Virginia on November 20, 2023, shows a smart phone screen displaying the logo of OpenAI juxtaposed with a screen showing a photo of former OpenAI CEO Sam Altman attending the Asia-Pacific Economic Cooperation (APEC) Leaders' Week in San Francisco, California, on November 16, 2023.

OpenAI's announcement on Tuesday, May 28, introduced the Safety and Security Committee, which will oversee the company's safety practices and address the risks associated with its AI models, Reuters reports.

Alongside CEO Sam Altman, the committee includes directors Bret Taylor, Adam D'Angelo, and Nicole Seligman. Technical and policy experts Aleksander Madry, Lilian Weng, and John Schulman are also part of the team, along with newly appointed Chief Scientist Jakub Pachocki and head of security Matt Knight.

The primary task of this committee over the next 90 days is to evaluate and develop OpenAI's existing safety protocols. After this period, the committee will share its recommendations with the board, which will then publicly disclose the adopted measures.

Recent Leadership Changes at OpenAI

The formation of the new committee follows major leadership changes within OpenAI.

Earlier this month, Ilya Sutskever, the former Chief Scientist and co-leader of the disbanded Superalignment team, left the company.

The Superalignment team, established to ensure AI models align with intended objectives, was also dissolved in May, and its members were reassigned to other projects.

Jan Leike, who also co-led the Superalignment team, resigned alongside Sutskever, raising concerns about OpenAI's approach to safety.

John Schulman, a co-founder of OpenAI who previously headed the team that developed ChatGPT, will now lead the company's long-term safety research. His work will be overseen by the new Safety and Security Committee, ensuring a cohesive strategy for managing AI risks.

OpenAI Now Training Next Flagship AI Model

In addition to forming the new safety committee, OpenAI has started training its next AI model, aimed at succeeding GPT-4.

As reported by the New York Times, this new model is expected to bring "the next level of capabilities," according to OpenAI. It represents a significant step towards developing artificial general intelligence (AGI), a machine that can perform any intellectual task that a human can.

The upcoming model is intended to power several of AI products, including chatbots, digital assistants, search engines, and image generators.

Reports tell us that the new model may not be available for another nine months to a year or more.

"While we are proud to build and release models that are industry-leading on both capabilities and safety, we welcome a robust debate at this important moment," the company stated in a recent blog post.

Recently, the company was accused of ripping off Hollywood actress Scarlett Johansson's voice to power ChatGPT's Sky voice.

Stay posted here at Tech Times.