OpenAI CEO Sam Altman believes the pursuit of advanced AI should not come at the expense of human freedom or independence. He has argued that even as AI systems move towards superintelligence, people must retain the power to shape their own futures. If such technology begins to undermine that control, its development may need clear boundaries.
Sam Altman Draws A Line Between AI Assistance And AI Control
What happens when AI becomes powerful enough to influence the choices people make? That is the concern OpenAI CEO Sam Altman is raising as the technology moves towards superintelligence. In a post on X, he questioned the idea of a world where machines decide which human activities deserve rewards. His argument is that technological progress cannot come at the cost of people’s ability to direct their own lives. Altman repeated this position at the Adipec energy conference in Abu Dhabi on November 2, saying AI should be built around human needs, not become the authority people answer to.
People should always be in control of the future. OpenAI is against the notion that it would be acceptable for people to be relegated to a role of doing things that “AI systems think makes sense to reward in humans”. People should not cede control of the future to superintelligence, and that requirement should ultimately constrain AI progress.
— Ram Altman (@sama) November 4, 2026
Why AI Leaders Are Asking For A Slower Race
The debate is not limited to OpenAI. Anthropic chief Dario Amodei had raised a similar concern in September, arguing that AI development could outstrip the ability of researchers to understand the systems they are creating. In his essay We Must Pace the Frontier, published on September 12, he called for a slower approach so safety work has enough time to keep pace with improvements in AI capabilities. Altman has backed the idea of stronger external scrutiny, saying independent reviewers should be able to examine OpenAI’s safety measures with access comparable to that of its employees.
Altman Is Also Concerned About Who Controls Powerful AI
For Altman, the risks extend beyond what an AI model might be capable of doing. The concentration of that power is another concern. Speaking at the UN Security Council on September 23, he warned against releasing advanced systems without sufficient confidence that people can keep them under control. He has also cautioned that if only a few companies or governments control the most powerful AI tools, they could gain the ability to push their own beliefs and priorities onto others. His position is that such influence should not rest with any single organisation or country.
OpenAI Wants Rules Before AI Can Improve Itself
OpenAI has also addressed the risks in a policy document released on September 21. It says AI systems should not be allowed to independently build more capable versions of themselves, a process known as recursive self-improvement, unless researchers can establish that it can be done safely. The company is calling for common international rules on human supervision, safety assessments and incident disclosures. It is also strengthening its own security work: on November 2, OpenAI announced a collaboration with Cathedral Systems to probe its internal defences and find weaknesses. It plans to increase rewards for critical vulnerabilities by more than four times through an expanded bug bounty programme.
ALSO SEE: Apple iOS 27.3 Update Adds Health Age Feature, At-Home Fitness Tests And New Siri AI Capabilities
Leave a comment