A rare united front on pacing

The core of the claim holds up. Leading AI executives and researchers have backed calls to slow the pace of development, including an open letter urging a temporary halt to training models more powerful than the current frontier, and a joint statement from industry chiefs flagging the risk of human extinction from AI as a global concern. The review, which spanned 11 primary sources and relied on search-page cross-checking — with verification hampered at points by bot-blocking walls — found the substance of the statements consistent with the claim.

Where the claim bends

The "one voice" framing is where nuance enters. The appeals are broad and voluntary: none binds any company to a timeline, checkpoint or external audit, and the same leaders urging caution continue to ship products, court investment and expand data-center buildouts. Within the research community, prominent voices also dispute the extinction framing itself, arguing it distracts from near-term harms. The claim's strongest reading — a fully unified industry demanding a slowdown — therefore overshoots what the record shows.

Stakes for Korea's chip and AI camp

For Seoul's ecosystem, the debate cuts both ways. Samsung Electronics and SK hynix sit on the supply side as makers of the semiconductors behind LLM training, while Naver and Kakao are developing consumer-facing models of their own. Any move toward global pacing agreements or AI safety regulation would land directly on both halves of that portfolio, from memory and foundry orders to model roadmaps — giving Korean players a front-row stake in whether the "slow down" chorus translates into policy.

The record supports the substance of the claim while trimming its superlatives: AI's most influential figures have indeed urged a slower pace amid extinction fears, though the appeals remain voluntary and contested — making the claim Mostly True.