Codenotary launches AI security platform that learns from AI agent behavior
2026-07-08T08:51:48Z•b709751a59982cb952d501c00ae9acb6be683c7ced4699f61ecf720418591bc0
AI agentsAWS MarketplaceAgentMonAnthropicBarracudaCVE exploitabilityClaude CoworkCodenotaryCyberProofEvo SecurityMXDROpenAI vs AnthropicOrbiaPAMPicusadaptive runtime securityagentic AIautonomous exposure validationbehavioral learningcontinuous offensive testingidentity securityindustrial cybersecuritymacOS agentsopen-source security tools
What happened
News roundup focused on AI-agent security and related tooling. Codenotary announced AgentMon 3, introducing adaptive runtime security policies that learn from customer-specific AI agent behavior and is now available on AWS Marketplace. Anthropic rolled out Claude Cowork for mobile/web to run multi-step, background and scheduled agent tasks; macOS is highlighted as a proving ground for unattended agents. Industry moves include Barracuda’s acquisition of Evo Security to add PAM and identity protection to BarracudaONE, CyberProof’s Agentic MXDR service combining expert human oversight with AI-ag-
Why it matters
A reviewed impact interpretation has not been published for this record.
Evidence and limitations
- Source ID
- helpnetsecurity
- Record identifier
- b709751a59982cb952d501c00ae9acb6be683c7ced4699f61ecf720418591bc0
- Enrichment time
- 2026-07-08T08:51:48Z
- AI-assisted enrichment
- Yes
This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.