Who pays when you gate cyber-capable AI models?

This interview with Jaya Baloo (COO & CISO at Aisle) evaluates arguments for restricting cyber-capable AI models, concluding that gating may slow attacker proliferation but risks harming defenders, widening capability gaps for smaller organizations and stressing the software supply chain; she recommends CISOs assume adversaries have access, invest in faster detection, stronger identity and supply-chain controls, and prioritize rapid remediation.