news
Scmp World
2026-08-04
US AI leaders suggest open-weight models may be safer than closed-source options
Prominent AI experts are challenging the narrative that open-weight models are inherently dangerous, arguing they may offer better safety and security than closed-source alternatives.