The Intent-Based Security Platform by Keystrike is a session-layer control system that provides continuous verification of human intent behind every post-authentication action across privileged remote sessions and non-interactive protocols. The platform captures live visibility into interactive protocols including RDP, SSH, VNC, TeamViewer, NinjaOne, SplashTop, and vendor connections, and extends governance to non-interactive protocols such as PowerShell Remoting, WinRM, PsExec/SMB, and WMI via a jump-host chokepoint. It verifies that each action originates from physical human input using a patent-pending device-bound cryptographic attestation mechanism with TPM-backed key storage to sign each hardware input event, enforcing deterministic, fail-closed blocking of commands that cannot be attested. The platform also extends verification and enforcement to AI-driven agents, offering agentless visibility into AI activity from managed endpoints and applying person-bound enforcement for agent actions by linking AI-agent actions to device-bound cryptographic evidence that traces back to a specific authorizing individual.
The platform generates tamper-evident, non-repudiable evidence for every governed action, including per-action signed attestations and logs of session lifecycle events, enabling audit-ready proof of activity without requiring agents on every downstream system. It can operate in both agent-based and agentless modes, allowing time-boxed temporary access for vendors without client installation, and records whether access originates from approved employee workstations or unmanaged devices. Evidence can be integrated via webhooks into existing security tools and aligned with compliance frameworks such as NIST CSF and NIS2. The system is designed to complement existing identity, remote-access, IAM, MFA, PAM, ZTNA, IGA, EDR, and SIEM solutions by adding a control point that ensures authorized identities translate into verified behavior during the session.
