What the claim asserts

The claim rests on three elements. First, that Altman has conceded that AI carries risks. Second, that he has nevertheless appealed for continued trust in companies — rendered in the claim as "still, trust the companies." Third, that the juxtaposition has fed a spreading argument over whether AI developers should be allowed to police themselves rather than submit to externally imposed rules. The claim circulated in Korean-language video coverage, and the review examined it as presented there.

What the review found

Reviewers worked through eight primary sources, and the record backed the essentials: Altman's acknowledgment of AI risk is on the record, and so is his appeal for continued confidence in the industry's own stewardship. The third element — a self-regulation debate in full spread — was likewise judged grounded rather than embellished; the question of whether developers should set their own guardrails for LLMs and the systems built on them, or be bound by outside rules, has genuinely intensified. The rating fell short of the top grade, however, indicating that while the substance stands, elements of the claim's framing or emphasis drew reservations.

Why it matters

The dispute is more than a quarrel over one executive's phrasing. If the industry's answer to AI risk is reassurance — trust the builders — then the credibility of that reassurance becomes a policy question in its own right, one that lawmakers, safety researchers and companies are now arguing openly. For hardware and software players outside the United States, including Korea's AI and semiconductor sector, the outcome will help define the supervisory landscape for years. On the evidence assembled, the verdict is Mostly True.