Major Garrett hosts The Takeout on Friday, September 17, and the show opens with two warnings. One comes from OpenAI. The other comes from the United Nations.
OpenAI has reported six new incidents of “concerning” behavior. A U.N. panel says it has “reasonable grounds” to believe the United States committed war crimes in Iran.
OpenAI’s Six New Incidents
The report flags behavior that OpenAI considers worrisome. The fact that six new cases were flagged suggests continued attention to how its systems operate. The report does not describe the nature of the incidents beyond labeling them as concerning.
The U.N. Panel on Iran
The U.N. panel’s statement is more explicit. It says it has reasonable grounds to believe the United States committed war crimes in Iran. Reasonable grounds signals that evidence has been gathered and evaluated, though it does not confirm guilt.
Two Warnings on One Page
The two stories sit side by side on the show’s agenda. One is technical and corporate. The other is political and international. They come from different fields, but they share a common structure: both warn of consequences that have not yet arrived but could.
Comparing the Two Warnings
| Item | What Happened |
|---|---|
| OpenAI | Reported six new incidents of “concerning” behavior |
| U.N. panel | Says it has “reasonable grounds” to believe U.S. committed war crimes in Iran |
Where the paper stands
The paper backs OpenAI’s disclosure of safety failures, even while it warns against broad regulation that would favor big tech over startups, and it is against the U.N. panel’s war-crimes claim being used as justification for the kind of heavy-handed oversight that would entrench the largest firms.
OpenAI reporting its own failures is the right path. A startup can’t compete with a giant if the giant gets to hide its faults. But the U.N. panel’s war-crimes claim should not become a lever for federal control that locks in today’s leaders. That would turn a warning into a moat, not a safeguard.
What readers should watch for is any move toward federal licensing or heavy compliance costs for AI. Those are the dangers, not the technology itself.
Key Facts Box
- Date: Friday, September 17
- Show: The Takeout with Major Garrett
- OpenAI: Six new incidents of concerning behavior reported
- U.N. panel: Reasonable grounds for belief of war crimes
- Location: Iran
The two warnings share a common thread. Both involve consequences that have not yet arrived but could. OpenAI’s warning concerns the future of its own systems; the U.N. panel’s concerns the future of a nation and its people.
OpenAI’s report is less dramatic in form but still consequential. Six new incidents of concerning behavior add to the company’s existing record. The fact that these incidents continue to occur suggests that the underlying problem has not been fully resolved.
The U.N. panel’s statement is significant. Reasonable grounds signals that evidence has been gathered and evaluated, though it does not confirm guilt.
The two stories together paint a picture of a world where technology and war are increasingly intertwined. The consequences of both are yet to be fully understood.
See the video the story is built around at CBS News.
Get the Notebook.
The day's best stories and every fresh verdict, in plain English, in your inbox by seven. One email a day, no more.