‘Too Dangerous to Release’ Is Becoming AI’s New Normal

CSET’s Steph Batalis shared her expert insight in an article published by TIME. The article examines how leading AI companies are increasingly restricting access to their most capable models, such as GPT-Rosalind and Claude Mythos, due to growing concerns around dual-use risks in areas like cybersecurity and biological research, and the broader question of who should govern access to these systems.