Sam Altman Warns AI Progress Is 'Scary' as OpenAI Slows Development Over Safety Concerns

Sam Altman Acknowledges Growing Unease Over AI Development

Sam Altman, CEO of OpenAI, has publicly stated that the rapid progress of AI is "scary" and unsettling, even for those directly involved in building the technology1

1

2

2

. In a recent post on X, Altman revealed that OpenAI has been navigating a complex tension between excitement about new capabilities and anxiety about where increasingly capable AI systems might lead. "We have been living with the tension between being excited and anxious about progress for some time," Altman said, adding that this discordance is "much more" pronounced for people outside the industry1

1

. His comments come as OpenAI prepares to introduce its next major model while facing mounting questions over AI safety, public trust, and the real-world impact of powerful AI systems.

OpenAI Slows Model Rollout to Prioritize Safety and Alignment

Altman revealed that OpenAI spent much of the summer focused on safety-related priorities and is now taking a more measured approach with future models2

2

. While expressing confidence in the company's latest work, he emphasized that OpenAI is deliberately being more cautious with models that follow to ensure adequate time for evaluating safety and alignment as AI systems become more capable2

2

. The company's next major model, Astra, has already completed training. Altman described it as "a big upgrade in both capability and alignment," stating that OpenAI is proud of what it has built while acknowledging that increasingly powerful AI warrants a more measured approach2

2

. This shift reflects OpenAI's recognition that safeguards must advance alongside technological progress to maintain public trust and ensure responsible deployment.

Long-Term Consequences Remain Unclear Despite Rapid Advancement

Altman issued a stark warning about the uncertainty surrounding advanced AI. "AI is getting extremely capable; no one fully understands the consequences of this," he said1

1

. He argued that many of the long-term consequences of increasingly capable AI systems remain misunderstood and that managing the transition toward widely available and powerful AI should become a global prioritization1

1

2

2

. However, Altman emphasized that slowing technological progress alone would not solve the problem. Instead, he believes society needs to understand how AI systems behave in real-world situations while developers improve safeguards1

1

.

Societal Adaptation to AI Becomes Critical Priority

Altman stressed that the challenge extends beyond what AI companies can address internally. He stated that society must adapt alongside increasingly powerful AI systems as their real-world impact grows2

2

. According to Altman, the world needs to understand how quickly AI is developing and what powerful models can mean in practice2

2

. He argued that keeping AI development aligned with human interest and strengthening safeguards will be critical as society moves into the next phase of technology2

2

. This call for societal adaptation to AI suggests that regulatory frameworks, educational initiatives, and public discourse around AI capabilities must accelerate to match the pace of technological advancement and ensure responsible integration into daily life.