Would you trust a critical system that nobody outside the vendor can audit?
That's exactly what closed models demand. Systems you can't inspect, test, or run yourself.
That's not security. That's dependence.
The cybersecurity debate on open-source AI is backwards. Open models aren't the risk, they're the defense! Attackers can already jailbreak any API or guardrails. Defenders can't secure systems with black boxes they can't control, inspect, test, or run locally.