What happened
Cisco announced AI Defense 2.0 on August 6, extending its security platform to monitor enterprise AI agents that call external tools and APIs. The product inspects agent prompts, tool permissions and data exfiltration patterns across Slack, Microsoft 365 and custom agent gateways.
Cisco said early customers include three Fortune 100 financial firms running pilot deployments. AI Defense integrates with existing Cisco Secure Access and Splunk logging pipelines.
Why it matters
Security teams are shifting from blocking chatbots to governing agents that can act on systems. Tool-access monitoring is a new control plane that traditional DLP and CASB products were not built for.
Financial services pilots suggest regulated industries will demand agent-specific security before wide rollout.
MintedBrain take
If agents can reach production APIs, add tool-call logging before adding blocking rules. You need baseline behavior data to tune policies without breaking legitimate workflows.
Discussion
Sign in to comment. Your account must be at least 1 day old.