Future of Life Institute

Releasing full model weights may allow malicious actors to strip or override safety mechanisms, creating uncensored or harmful versions. In contrast, supervised fine-tuning preserves core safety guardrails while enabling responsible customization.

Leading AI companies should: Ensure Human-Legible Reasoning - AI models should reason in ways that are accessible and understandable to humans. Developers should avoid opaque reasoning methods.

Strengthen whistleblower protections and align culture with policy.

Proactive incident reporting enables collective learning from safety failures and near-misses.

Broaden dangerous-capability evaluations and link thresholds to binding mitigations. Evaluations have gaping holes (no AI R&D data), and no procedure connects threshold breaches to deployment decisions, making thresholds effectively non-binding.

The initiative calls for a prohibition on the development of superintelligence until the technology is reliably safe and controllable, and has public buy-in.

provisions to outlaw AI personhood to ensure AI companies and malign actors remain responsible for their actions.

We commend the BIS rule’s inclusion of a provision which requires the reporting [...] computing power greater than 10^{26} integer or floating-point operations.