OpenAI told staff this week that it has deliberately slowed parts of its model development and paused specific internal AI‑training runs, while urging other labs to adopt a voluntary industry‑wide slowdown until common safety standards are established.
Internal pauses and a possible slowdown
At a company‑wide meeting, CEO Sam Altman said OpenAI may "drosseln" (slow down) the pace of its AI development, potentially in coordination with other laboratories. The announcement was reported by Handelsblatt and attributed to internal sources cited by Bloomberg.
"Bei einer unternehmensweiten Versammlung sagte Vorstandschef Sam Altman in dieser Woche, OpenAI könne das Tempo seiner KI‑Entwicklung möglicherweise drosseln – eventuell gemeinsam mit mehreren anderen KI‑Laboren."
In addition to the broader slowdown discussion, OpenAI disclosed that it has already paused certain internal training runs for safety reasons. The pause applies to "bestimmte interne KI‑Trainings" that were deemed to pose heightened risk, though the company did not specify which models or projects were affected.
Call for a voluntary industry slowdown
Jakub Pachocki, OpenAI’s chief scientist, used a recent blog post to warn that AI firms should coordinate their development speed. He wrote that companies should "nach Bedarf abstimmen" (coordinate as needed) to slow future progress until shared safety standards are in place.
"Er sei der Ansicht, dass sich KI‑Unternehmen 'nach Bedarf abstimmen sollten, um die zukünftige Entwicklung zu verlangsamen'."
Pachocki added that he hopes "freiwillige Verlangsamungen üblich werden, bis gemeinsame Sicherheitsstandards etabliert sind" – a direct appeal for a sector‑wide, voluntary pause.
Policy backdrop and White House talks
Altman told staff that he discussed the need for a slowdown with White House officials in July. The conversation, also reported by Handelsblatt, underscores the growing governmental interest in AI safety and the possibility of future regulation.
OpenAI has not issued a formal comment beyond the internal briefing, and the company’s public blog or press releases have not yet addressed the slowdown. The lack of a public statement leaves investors and partners without clear guidance on how the pause will affect product roadmaps or revenue forecasts.
Company background
OpenAI is a San Francisco‑based artificial‑intelligence firm founded on 11 December 2015. The organization employs roughly 4,500 staff members, according to its Wikidata entry. Sam Altman serves as chief executive. While the Wikidata record provides these baseline facts, the research packet notes that the figures may lag reality and should be cross‑checked against the company’s own disclosures before publication.
What remains unknown
- The specific models or projects affected by the internal training pause have not been identified.
- OpenAI has not disclosed how long the pause will last or whether it will impact upcoming product releases.
- Other AI labs have not publicly responded to the voluntary slowdown appeal, leaving the sector’s collective commitment uncertain.
- Regulatory outcomes from the July White House discussion are still unclear.
Analysts will watch OpenAI’s next earnings call and any follow‑up statements for clues on whether the slowdown will become a lasting policy or remain a temporary safety measure.