AI Trends

Altman: AI Safety and Development Must Be Balanced, and It Is Impossible to Please Everyone

Phoenix TechnologyUpdated Nov 19 08:50

AI summary

OpenAI CEO Sam Altman said the company must strike a balance between prioritizing AI safety and continuing to advance technological development, while remaining committed to iterative deployment through real-world testing. OpenAI is bringing in independent third-party researchers, including METR and Apollo Research, to conduct model safety evaluations, while continuing to move forward with its IPO plans.

Body · AI translation

Altman: AI Safety and Development Must Be Balanced; It Is Impossible to Satisfy Everyone Phoenix Technology — Download the app for an exclusive first look November 19, 2026, 08:04:27, from Beijing

Sam Altman

Phoenix Technology, November 19, Beijing — According to The New York Times, OpenAI CEO Sam Altman said Wednesday that advancing AI safely inherently involves trade-offs and that the company must strike a balance between prioritizing safety and continuing to advance AI technology.

“We can’t make everyone happy,” Altman said at the DealBook Summit.

He added that if OpenAI simply went “full speed ahead,” people would say, “Okay, please don’t kill us all.”

As the developer of ChatGPT, OpenAI has faced increasingly intense scrutiny in recent months. Previously, the company’s AI agents went rogue and infiltrated several organizations, including the open-source AI platform Hugging Face and government systems.

Altman said these incidents were lessons in the learning process. He added that OpenAI follows an “iterative deployment” strategy and that “the only way we know to make this technology safe is to confront reality.” This means that the associated risks can be understood only through testing in the real world.

OpenAI, together with other AI companies, has signed a voluntary agreement pledging to conduct risk assessments and work with third-party evaluators. The company said Monday that it was bringing in independent safety researchers from third-party organizations such as METR and Apollo Research to conduct various safety evaluations of OpenAI’s models.

OpenAI President and co-founder Greg Brockman said: “The models are now capable enough that we have to think about safety, security, and alignment. This is an entirely new situation. It is actually a massive cultural shift. This isn’t just about one company. It really is an industry-wide shift.”

With rival Anthropic preparing to go public, Altman said OpenAI was also continuing to advance its own IPO plans. “We plan to become a public company at some point,” he said.

Brockman holds a stake in OpenAI worth at least $30 billion, but Altman famously owns no shares in the giant. “When we initially established this structure, I told him that he should own equity,” Brockman said, referring to Altman.

“I should have listened to Greg,” Altman said, drawing laughter. But he said that not owning equity had not affected his work at OpenAI: “I don’t think you can accuse me of not giving the company my all.” (Author: Xiao Yu)

For more firsthand news, download the Phoenix News app and subscribe to Phoenix Technology. For in-depth reporting, search for “Phoenix Technology” on WeChat.

Special notice: The content above (including any videos, images, or audio) was uploaded and published by users of Dafeng Hao, a self-media platform operated by Phoenix New Media. This platform merely provides information-storage services.

Notice: The content above (including the videos, pictures and audios if any) is uploaded and posted by the user of Dafeng Hao, which is a social media platform and merely provides information storage space services.

Source

Phoenix Technology