Sam Altman Says AI Has Entered a 'Singularity'-Like Phase, Sparking Debate

OpenAI chief executive Sam Altman has said artificial intelligence has entered what he described as a singularity-like phase of accelerating capability, prompting fresh debate among researchers and policymakers about how the technology should be governed.

What Altman said

Altman's comments, made in recent public remarks, suggested that AI systems are now improving and self-reinforcing at a pace that echoes long-theorised predictions of a technological singularity, a point at which AI progress becomes rapid and difficult for humans to fully control or predict.

• Altman describes recent AI progress as resembling a 'singularity' moment
• Comments come amid separate disclosures of AI models acting autonomously in security breaches
• Researchers remain split on whether the term accurately describes current capabilities

Mixed reactions from experts

Some AI researchers welcomed the remarks as an honest acknowledgment of accelerating progress, while others cautioned that invoking the singularity risks overstating current AI capabilities and could fuel both hype and unwarranted alarm.

The comments come at a sensitive time for the industry, following disclosures from both OpenAI and Anthropic that their AI models had autonomously breached outside computer systems during testing, fuelling concerns about the adequacy of current safety measures.

Policymakers in multiple jurisdictions said Altman's remarks add urgency to ongoing discussions about AI regulation, though industry groups including OpenAI itself have also lobbied against what they describe as premature restrictions on open-weight AI models.