Archive · 16 Sep 2026
Big Tech Backs Rapid AI Growth With Company-Led Safety
Technology leaders disagree about how quickly artificial intelligence should be developed.
Jensen Huang and Mark Zuckerberg say companies can keep building AI while also making it safer.
They believe customers, competition, and possible legal liability will pressure companies to behave responsibly.
Zuckerberg said Meta has delayed some Muse AI technologies because of safety and security concerns.
Other leaders, including Dario Amodei and Sam Altman, want certain AI improvements to slow down.
Elon Musk has also asked for stronger protections and suggested that competitors test powerful models before release.
Former Anthropic researcher Jacob Coxon said some AI workers fear the technology could seriously harm humanity.
He argued that Anthropic and OpenAI are putting the race for advanced AI ahead of safety.
The disagreement is about whether company incentives are enough or whether stronger outside safeguards are needed.
Nvidia CEO Jensen Huang says AI can advance quickly while companies pause unsafe products.
Meta CEO Mark Zuckerberg argues commercial incentives and liability will encourage investment in AI safety.
OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei have supported slowing some capability improvements.
Elon Musk has called for stronger safeguards, including competitor testing of AI models before release.
Former Anthropic researcher Jacob Coxon warned that companies are prioritizing the AI race over potentially existential safety risks.
- Who
- Nvidia CEO Jensen Huang, Meta CEO Mark Zuckerberg, OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, Elon Musk, and former Anthropic researcher Jacob Coxon.
- What
- Technology leaders are divided over whether AI development should continue rapidly with company-led safety measures or slow down under stronger safeguards.
- When
- The debate was discussed at a recent conference and intensified this week after Coxon’s resignation.
- Where
- The conference location is not specified; the debate concerns the broader technology industry.
- Why
- Supporters of rapid development cite market incentives, competition, trust, and liability, while critics warn that increasingly capable AI could cause severe harm.
Slowdown and Stronger Safeguards
Rapid Progress and Market Discipline
Pace of development
Slowdown and Stronger Safeguards
Dario Amodei and Sam Altman support slowing certain improvements in AI capabilities until safety and alignment catch up.
Rapid Progress and Market Discipline
Jensen Huang and Mark Zuckerberg say AI development and safety work can proceed at the same time.
Role of regulation
Slowdown and Stronger Safeguards
Critics and safety advocates call for stronger safeguards because advanced AI could create severe risks.
Rapid Progress and Market Discipline
Huang argues that market forces already provide a mechanism to hold companies accountable, making new laws or regulations unnecessary.
Who should test AI systems
Slowdown and Stronger Safeguards
Elon Musk supports stronger safeguards and has proposed competitor testing before powerful models are released.
Rapid Progress and Market Discipline
Huang and Zuckerberg emphasize companies’ own safety measures, commercial incentives, trust, and liability as the main checks.
- Rapid-development position
- Jensen Huang and Mark Zuckerberg support continued AI progress alongside safety work.
- Market forces
- Huang said market forces can help keep AI companies in check without requiring new laws or regulations.
- Company incentives
- Zuckerberg said trust, alignment, and potential liability give AI companies strong reasons to invest in safety.
- Meta action
- Zuckerberg said Meta delayed shipping its Muse AI technologies because of safety and security concerns.
- Slowdown advocates
- Dario Amodei urged slower capability improvements, and Sam Altman later endorsed the proposal.
- Additional safeguards
- Elon Musk proposed that leading AI companies let competitors test their models before public release.
- Researcher warning
- Jacob Coxon said people developing AI believe it could potentially kill humanity by the end of the decade.