WitnessAI · Security requirement
Agent-to-agent communication security
Research note
Superseded stale absence record for WitnessAI agent-to-agent security.
Public-source research
Vendor evidence
Review what vendors say publicly, the exact quoted source text, the security requirement each statement may support, and what still needs verification.
Research library coverageCounts describe the research workflow, not vendor quality or product effectiveness.View details
Evidence records1280
Source checked863
Needs verification0
Source captured0
No supporting claim found411
Excluded from evidence6
Showing 681–700 of 1280 evidence records
WitnessAI · Security requirement
Non-human identity and service-account security
What the vendor says
WitnessAI claims governance across human and AI-agent workforces, including visibility into agent tools and data access.
Exact source quote
human and digital workforce
Lasso Security · Security requirement
Unapproved AI use discovery
What the vendor says
Lasso Security claims it discovers AI agents and homegrown AI applications via platform integration and CI.
Exact source quote
Discover AI agents via platform integration. Connect via CI to automatically discover homegrown applications.
Lasso Security · Security requirement
AI-feature discovery in business applications
Research finding
No public claim found for this capability.
Lasso Security · Security requirement
Approved AI usage monitoring
Research finding
No public claim found for this capability.
Lasso Security · Security requirement
Controls for unapproved AI use
What the vendor says
Lasso Security claims runtime policy enforcement through proxy, application programming interface (API), or AI gateway controls.
Exact source quote
Enforce policies inline at the proxy, API, or AI Gateway layer with runtime protection that adapts as your application evolves.
Lasso Security · Security requirement
Sensitive-data protection for generative AI
What the vendor says
Lasso Security claims it analyzes what sensitive data agents can reach and enforces policies inline at the proxy, application programming interface (API), or AI Gateway layer.
Exact source quote
Analyze your security posture by understanding what your agents are exposed to, what actions they can perform, what sensitive data can they reach, and more.
Lasso Security · Security requirement
Action-taking agent monitoring
What the vendor says
Lasso Security claims it maps models, system prompts, tools, and guardrails and provides full context on attacks including which agent was targeted.
Exact source quote
Maps models, system prompts, tools and guardrails
Lasso Security · Security requirement
Agent-to-agent communication security
What the vendor says
Lasso Security claims AI detection and response coverage for agent threats, including tool-chain manipulation and abnormal AI behavior.
Exact source quote
what actions they can perform, what sensitive data can they reach
Lasso Security · Security requirement
Non-human identity and service-account security
What the vendor says
Lasso Security claims AI agent inventory includes exposed actions and sensitive data reachability to support agent risk analysis.
Exact source quote
what actions they can perform, what sensitive data can they reach
Harmonic Security · Security requirement
Unapproved AI use discovery
What the vendor says
Harmonic claims visibility across the full AI stack across vendors, including long-tail and agentic surfaces, beyond Microsoft-only coverage.
Exact source quote
Purview gives you visibility inside Microsoft, on Microsoft tools, with Microsoft pattern matching. Real AI usage is not Microsoft-only. We see the full stack across vendors, including the long tail and the agentic surfaces, and we govern with intent classification rather than regex.
Harmonic Security · Security requirement
AI-feature discovery in business applications
What the vendor says
Harmonic claims coverage for AI running in browser tabs, desktop apps, employee-started agents, and embedded copilots inside software as a service (SaaS) tools.
Exact source quote
AI does not live in one place. It runs in the browser tab your sales lead opened, the desktop app your developer installed, the agent your engineer kicked off, and the embedded copilot inside the SaaS tools you already pay for. Harmonic Security covers all four.
Harmonic Security · Security requirement
Approved AI usage monitoring
What the vendor says
Harmonic claims it understands intent behind AI interactions across approved tools, shadow apps, and employee-created agents and governs those interactions in real time.
Exact source quote
AI is everywhere your employees work: in approved tools, shadow apps, and the agents they're spinning up on their own. Harmonic Security sits at that layer, understands the intent behind every interaction, and governs it in real time.
Harmonic Security · Security requirement
Controls for unapproved AI use
What the vendor says
Harmonic claims teams can block risky AI actions in real time, warn employees with context, or log silently for review.
Exact source quote
You can block in real time, warn the employee with context about why the action is risky, or log silently for security team review.
Harmonic Security · Security requirement
Sensitive-data protection for generative AI
What the vendor says
Harmonic claims data classification and logging provide audit trails, data residency controls, and evidence that AI use stays within defined boundaries.
Exact source quote
Our data classification and logging give you the audit trail, the data residency controls, and the ability to demonstrate that AI use in your organization operates within defined boundaries.
Harmonic Security · Security requirement
Action-taking agent monitoring
What the vendor says
Harmonic Security claims it governs AI interactions across employees and agents, including browser, desktop, agent, and Model Context Protocol (MCP) surfaces.
Exact source quote
Every interaction visible. Every interaction governable.
Harmonic Security · Security requirement
Agent-to-agent communication security
What the vendor says
Harmonic Security claims coverage for agents, skills, Model Context Protocol (MCP), and CLI workflows, including tool calls, scopes, and destructive actions.
Exact source quote
Tool calls, scopes, destructive actions.
Harmonic Security · Security requirement
Non-human identity and service-account security
Research finding
No public claim found for this capability.
Reco · Security requirement
Unapproved AI use discovery
What the vendor says
Reco claims it automatically discovers AI agents across Copilot, ChatGPT, Claude, Agentforce, Make, n8n, and custom integrations.
Exact source quote
Complete Inventory Automatically discover every agent across Copilot, ChatGPT, Claude, Agentforce, Make, n8n, and custom integrations.
Reco · Security requirement
AI-feature discovery in business applications
What the vendor says
Reco claims embedded AI in software as a service (SaaS) apps, generative AI tools, and copilots is expanding faster than teams can track and connects to enterprise data.
Exact source quote
Embedded AI in your SaaS apps, GenAI tools, and copilots is expanding faster than your team can track. Every new feature connects to your data.