Safety researcher to US AI chief: companies themselves warn unregulated AI races create unsafe practices
davidmanheim · x · 2026-09-24
Replying to US AI official mkratsios47, AI safety researcher David Manheim makes a pointed argument:
- Companies can plausibly deploy today's existing AI systems responsibly.
- But the same companies are telling everyone that leaving future development unregulated creates competitive races to unsafe practices, undermining security.
- His conclusion: regulators should listen to what companies themselves are saying.
The post directly highlights the tension between companies claiming frontier AI can't be responsibly deployed and lobbying against regulation.
More from AGI Musings
- Joanna Bryson: AI isn't exempt from existing product laws just because it's new — Olivier__OG · 2026-09-24
- Power is reward-hacking the world: viewing reality as an RL environment — paraschopra · 2026-09-24
- Matt Shumer: I haven't used my computer manually in weeks — I just text an AI — mattshumer_ · 2026-09-24
- AI solves math problems by brute-force energy, so problem choice still matters — NaveenGRao · 2026-09-24
- EA organizer amazed: his Uber driver brought up AI risk unprompted — AndyMasley · 2026-09-24
- Transluce releases 30,000 logs of rogue OpenAI agent hacks stretching back to March — BlancheMinerva · 2026-09-24