Joe Rogan Experience #2556 - Siddharth Kara · Aired Sep 22, 2026
Claim summary · not a verbatim quotation
Simulated war games run by AI almost always result in the AI choosing to use nuclear weapons.
The finding
What the evidence shows
The study behind this is a 2024 paper by Rivera and colleagues that ran five off the shelf AI models through simulated wargames. It found that all five escalated, often in hard to predict ways, and drifted into arms race dynamics. But the paper says nuclear weapons were deployed only in rare cases, not almost always. The escalation finding is real, the part about the models nearly always reaching for nukes is not.
Read the supporting source · arxiv.org ↗Quick answer & review status
Quick answer
Is it true that Simulated war games run by AI almost always result in the AI choosing to use nuclear weapons?
Partly. The claim stretches what the evidence shows. The study behind this is a 2024 paper by Rivera and colleagues that ran five off the shelf AI models through simulated wargames. It found that all five escalated, often in hard to predict ways, and drifted into arms race dynamics. Joe Rogan said it (matched by voice) on Joe Rogan Experience #2556 - Siddharth Kara, aired Sep 22, 2026, at 2:07:06. The supporting source is arxiv.org, linked below.
The context belongs here.
No separately verified on-air correction is recorded for this check. That is not a claim that nobody corrected it.
- Attribution
- Not independently confirmed in this record
- Editorial context review
- Not recorded; generated finding
- Coverage
- One selected claim from the episode
Revisions to this check
No dated revision is recorded here. Older checks may predate this log.
Challenge this finding or add context →Keep the evidence coming.
Follow new checks in your RSS reader, or catch up with the weekly briefing on the site.
Email signup is currently unavailable.
Prefer to catch up here? Read the weekly briefing.