# Verity > The operating system for autonomous security research. Verity is an > authorized autonomous offensive security and exposure validation platform: > an AI red team, bug-bounty operations center, security-research laboratory, > and evidence vault in one governed system. A company connects the applications, APIs, cloud assets, repositories, and security programs it is authorized to test. Verity establishes exactly what it may test, researches targets with multiple AI agents, attempts real exploitation only through bounded, authorization-bound transports, and accepts a finding only after independent verification. A model saying "this looks vulnerable" is never treated as a finding. Key facts: - Machine-enforced authorization: verified ownership, normalized scope, permitted identities and techniques, request ceilings, and an expiring authorization grant. Missing, expired, revoked, or changed authority fails closed. - Independent verification: target-reaching proof, safe-sibling negative controls, critic review, replay, duplicate checks, and scope verification. Unprovable hypotheses are discarded. - Defensible evidence: actions are bound to authorization, identity, source, policy, scope, execution receipts, timestamps, SHA-256 digests, and custody references. - Full finding lifecycle: discovery, evidence, verification, human approval, reporting, remediation, and authorization-bound retest; for operator-run bounty work, also provider submission, dialogue, and outcomes. - Two isolated principals: owned_asset (customer assets and authorization) and bounty_program (operator-managed third-party programs). Neither inherits the other's credentials, authority, data, or evidence. - Provider ecosystems: Verity models 17 bug-bounty provider ecosystems; each provider operation stays disabled until its exact authority is established, and provider state comes only from authenticated provider evidence. - Detection scope today: six behavioral HTTP finders (BOLA (IDOR), reflected XSS, error-based SQL injection, template injection, path traversal, JWT token confusion); authorization bypass and SSRF are source-assisted. The source_audit lane is opt-in and requires verified Linux KVM gVisor isolation. - Reference-only secret handling: secret-manager references; raw credentials never enter product records. - Verity Inbound Reports: intake for reports from external researchers, customers, partners, and internal teams. - Not a vulnerability scanner, a Burp Suite wrapper, a generic LLM security chatbot, a bug-bounty marketplace, a breach-and-attack simulation catalog, a pentest report generator, or an agent allowed to attack anything it finds. ## Pages - [Platform overview](https://verity180.com/): positioning, capabilities, finding lifecycle, two-principal isolation, detection scope, and FAQ - [Product](https://verity180.com/product): platform architecture, research lanes, evidence review, patch verification, and reports - [Security](https://verity180.com/security): fail-closed execution controls, principal isolation, secret handling, and evidence manifests - [Compare](https://verity180.com/compare): Verity compared with XBOW, Terra Security, Horizon3.ai NodeZero, Pentera, SafeBreach, Picus, Cymulate, HackerOne, Bugcrowd, and Intigriti - [Controlled pilot](https://verity180.com/controlled-pilot): one application, two test identities, and one remediation cycle under explicit authorization ## Full text - [llms-full.txt](https://verity180.com/llms-full.txt): every capability, control, lifecycle step, FAQ answer, and comparison on the site as plain text ## Optional - [Sign in](https://verity180.com/login): customer access to the Owned-Asset Hunter console and Inbound Reports - [Inbound Reports intake](https://intake.verity180.com/): public evaluation and contact form for Verity Inbound Reports - [Sitemap](https://verity180.com/sitemap.xml) Contact: pilot@verity180.com