Sam Altman Says AI Benefits Are Worth Accepting Some Risks

OpenAI CEO Sam Altman says society may need to accept some negative consequences from artificial intelligence if it wants to preserve broad access to the technology.

He argues that AI’s overall benefits will outweigh the risks and that restricting powerful systems to a small number of companies would create an unacceptable trade-off.

Altman says broad AI access matters more than eliminating every risk

Altman said society should tolerate some risks from AI because keeping the technology broadly available could create significantly greater benefits, as Reuters reports.

Those risks could include hacking, scams, fraud, and other forms of misuse. However, Altman believes people will use AI for beneficial purposes far more often than harmful ones.

He also rejected the idea that only a handful of major AI laboratories should control access to increasingly capable systems.

Altman described that approach as “completely unacceptable,” arguing that preventing every possible misuse should not come at the cost of limiting public access.

OpenAI and Anthropic differ on AI regulation

Altman also highlighted what he described as a fundamental difference between OpenAI and Anthropic over how governments should regulate increasingly powerful AI systems.

OpenAI generally supports a lighter regulatory approach that allows companies to continue developing and distributing AI tools while addressing specific risks.

Anthropic has taken a more cautious position. CEO Dario Amodei has called for slowing development enough to “pace the frontier” as AI capabilities increase.

Altman has supported parts of Amodei’s argument, but the two companies continue to differ over how much regulation the industry needs.

Anthropic has also placed AI safety at the center of its messaging. The company recently dedicated about 80 pages of its IPO prospectus to potential AI risks, including threats to humanity, as covered in our report on Anthropic’s AI risk disclosures.

AI safety debate continues to grow

Altman’s comments come as governments and AI companies continue debating how to balance innovation, public access, safety, and regulation.

The discussion has become more prominent as AI systems gain additional capabilities and companies expand how they monetize and deploy them.

OpenAI is also making broader changes across its products. The company recently confirmed that ChatGPT is getting visual ads during image generation.

It also released GPT-6.1 Sol with near-Astra performance, while reports indicate that OpenAI scrapped the GPT-6.1 Astra release over safety concerns.