Comment by The 2026 Singapore Consensus on Global AI Safety Research Priorities

Priorities include: evaluations that still work even when a model can determine if it is being tested and may therefore understate its harmful capabilities (“sandbagging”) or conceal misalignment (“alignment faking”), backed by secure access for inde...

According to common benchmarks, open weight models are estimated to be 3 to 12 months behind the frontier (varying across capabilities). Developers of frontier closed models have recently designated these closed models as ‘High’ risk for cyber and bi...

Safety for open-weight models whose capabilities are nearing frontier closed models requires unique tools for open-weight model safety and ecosystem monitoring. [...] Accelerated public and private investment in research is required to keep pace with...