Mark Zuckerberg Says AI Will Not Kill Us 'If We Do a Good Job and Act Responsibly' in Interview

Mark Zuckerberg has said Artificial Intelligence (AI) will not kill us if everyone involved 'does a good job and acts responsibly', answering a direct question from tech journalist Joanna Stern about whether AI could kill us The Meta chief also said he is genuinely optimistic that AI will lead to a positive future for everybody, according to Hollywood Unlocked.

The comments were reported by Hollywood Unlocked on 26 September, after growing concern about the direction of AI development. The report said the discussion followed a former Anthropic researcher leaving the company over fears about where the technology is headed, while executives at rival AI labs discussed slowing development, adding guardrails and potentially involving governments more closely.

Mark Zuckerberg Gives His Take on AI Safety

Stern asked Zuckerberg directly, 'Is AI going to kill us?' According to Hollywood Unlocked, he laughed before answering that if everybody does a good job and acts responsibly, then 'no, AI isn't going to kill us'.

Zuckerberg also expressed optimism about the technology's future, saying he believes AI can become part of a positive future for everybody. He does not believe the industry needs to coordinate collectively to prevent problems, according to the report.

That position puts the emphasis on individual companies making responsible decisions as they develop AI. Zuckerberg pointed to Meta as an example, saying the company did not wait for others to give it the green light before making decisions it believed were right for both people and the business.

The answer is conditional, though. Zuckerberg's reassurance depends on the people developing AI doing their part responsibly, rather than treating the technology itself as automatically harmless.

Meta Takes a Different Approach to AI Development

Hollywood Unlocked also reported a disagreement over how quickly AI should advance. The report cited Anthropic's Dario Amodei among rivals pushing for the industry to slow the pace at which AI improves itself, describing it as AI effectively teaching itself to become smarter, faster.

Zuckerberg's position is different. He said Meta has already put the bulk of its computing power towards building technology that helps people in the present, rather than concentrating on systems that continue improving themselves.

He also argued that Meta did not need to wait for an industry-wide agreement before acting. The company, he said, made moves when it believed they were the right decisions.

Zuckerberg Points to Outside AI Evaluators

Zuckerberg also highlighted Meta's use of outside evaluators to examine the company's work. He pointed to those external checks as part of Meta's approach to AI safety and indicated that he believes every AI lab should be doing the same.

That adds another element to his argument. Individual companies can make their own decisions about AI development while taking steps to examine the technology they are building.

For Zuckerberg, the broader question is therefore less about stopping AI development altogether and more about whether the people building these systems will act responsibly. His answer to Stern was optimistic, but it came with a clear condition: technology needs the people behind it to get things right.