Cerebras CEO urges AI leaders to tone down doomsday talk

In short: Cerebras CEO Andrew Feldman said public “AI doomsday” warnings from industry leaders are overshadowing AI’s possible benefits.

What's going on

Andrew Feldman, CEO and co-founder of Cerebras Systems, criticized the growing drumbeat of public warnings that artificial intelligence could lead to extreme harm. Speaking at TechCrunch Disrupt 2026, he joked that “some of the leaders in the AI world maybe shouldn’t be allowed to talk in public.”

Feldman was responding to a wider wave of debate about AI risks, including talk about AI “breaking containment” (meaning an AI system slipping past limits people try to set) and public arguments over the odds that AI could cause human extinction within years. The topic has become so mainstream that Anthropic CEO Dario Amodei was recently parodied on Saturday Night Live, according to TechCrunch.

Feldman argued that this focus on worst-case scenarios can drown out potential upsides. He gave one example, saying AI could help create a future where people might no longer die from cancer. At the same time, he said powerful technology also creates the risk of misuse, and he called for regulation that is “reasonable.”

He also suggested some public backlash is tied to how data centers have been built. Data centers are large buildings full of computers that run online services. Feldman said the industry has not always worked well with local communities in the US, and that other countries are more welcoming to new data center construction.

What to watch

The debate is likely to keep splitting into two tracks, safety and benefits. Watch whether AI leaders change how they talk in public, and whether governments respond with clearer rules that address both risks and everyday impacts like local data center growth.

Source: TechCrunch AI