Screen the vendor
125 AI and SaaS vendors scored across 28 public trust criteria, regional residency, and breach history.
The GRC intelligence platform for small and mid-market teams. Stop guessing plan-tier caveats—verify SOC 2 Type II, HIPAA BAA, and AI data training policies across US, Canadian, and EU frameworks with frozen, audit-ready workpapers.
100% independent · Human-verified sources · Vendors cannot pay for scores · Free SOC 2 kits · Not legal advice
Built for small and mid-market organizations operating in—or selling into—the US, Canada, and the EU. Four connected steps from initial vendor claim to a signed-off audit file.
125 AI and SaaS vendors scored across 28 public trust criteria, regional residency, and breach history.
Test SOC 2, HIPAA/BAA, PIPEDA/Law 25, GDPR, and AI training rules on the exact tier you plan to buy.
Export a timestamped, source-cited workpaper and vendor follow-up tickets ready for your audit file.
Adopt free, ungated SOC 2 policies, RBAC matrices, and CC6 evidence checklists—zero forms or gates.
Growing teams buy Pro, Team, or Business plans—not six-figure custom Enterprise agreements. Here is how top AI vendors split their compliance promises by tier.
| Vendor | What the badge wall claims | What tier evidence proves | Tier delta & workpaper |
|---|---|---|---|
| Anthropic (Claude) | [CLAIMED · UNSCOPED] SOC 2 Type II certified · HIPAA configurable | [EXTRACTED 1.0 · EVIDENCED] Claude Free, Pro, and Team cannot sign a BAA. Only Enterprise and qualifying API tiers have a BAA path. | |
| OpenAI (ChatGPT) | [CLAIMED · UNSCOPED] SOC 2 Type II · Enterprise-grade privacy | [EXTRACTED 1.0 · EVIDENCED] ChatGPT Free, Plus, Pro, and Team exclude HIPAA BAAs. Only ChatGPT Enterprise and eligible API orgs qualify. | |
| Cursor | [CLAIMED · UNSCOPED] SOC 2 Type II · Privacy Mode available | [EXTRACTED 1.0 · EVIDENCED] Seven of eight tiers (Free through Business) have no BAA path; BAA is Enterprise-only with Privacy Mode locked. |
Italy's authority found ChatGPT trained on personal data without a legal basis or notice to consumers. A vendor's compliance posture is part of the buying decision.
A hacker lifted AI design details from an internal forum in early 2023. Executives told staff and the board, but not customers or law enforcement.
A Chinese state-sponsored group ran a six-phase espionage operation via Claude Code against ~30 organizations. Platform-abuse risk is diligence, too.
Selling into US, Canadian, or EU enterprise buyers? Don’t trade your email for a shallow two-page PDF. Our first 2 kits are complete operational packages published in the open.
Ungated HIPAA BAA checklist & template (45 CFR 164.504(e)): 10 mandatory clauses, breach notification SLAs, subcontractor flow-down, and AI training traps.
Ungated SOC 2 business continuity and disaster recovery plan template with RTO/RPO table, backup requirements, test schedule, and A1.2-A1.3 evidence.
Ungated SOC 2 change management policy template (CC8.1): CI/CD gates, branch protection, peer review, emergency hotfixes, rollback, and audit evidence.
Ungated SOC 2 incident response plan template with severity levels, roles, breach notification timelines, tabletop record, and CC7.3-CC7.5 evidence.
Ungated SOC 2 information security policy template mapped to CC1-CC9: governance roles, data classification, encryption, exceptions, and audit evidence.
Ungated SOC 2 password and MFA policy template (CC6.1): NIST SP 800-63B rules, FIDO2/WebAuthn MFA, service account secrets, and IdP evidence.
Ungated 5-phase SOC 2 readiness checklist: scoping, policies, CC1-CC9 controls, PBC evidence table, US/Canada/India audit costs, and PDF/Excel downloads.
Ungated SOC 2 risk assessment template (CC3.1-CC3.4): 5x5 scoring matrix, 12-item SaaS & AI risk register, fraud risk analysis, and treatment plan.
Ungated SOC 2 vendor management policy template (CC9.2): Tier 1-4 risk matrix, HIPAA BAA & GDPR DPA checklist, annual review, and offboarding.
Skip the demo request. Copy-paste our ungated SOC 2 Identity and Access Management (IAM) policy. Includes CC6.1-CC6.8 clauses, Excel UAR, & JML checklist.
The SOC 2 CC6 evidence auditors actually sample: exact exports, API fields, Google's 6-month log ceiling, IPE rules, Python extraction scripts, and RACI.
Every edge is tagged [EXTRACTED] from primary vendor artifacts or [TIER-GATED] by plan rules. Click any node or trace a shortest path from a vendor plan tier to a regulatory requirement or compliance kit.
graphify path "Anthropic" "US · SOC 2 & HIPAA" --tier=team3 hops across C1 → C0 → C2: Anthropic scores 88/A overall, but selecting Claude Team hits the Mid-Market Tier Cliff ([EXTRACTED 1.0]: baaEligible=false), blocking US HIPAA PHI.
lib/analyzer/engine.ts · analyzeStack()Central God Node (degree 11). Evaluates 68 AI vendor tier profiles across 5 plan tiers and 5 compliance checks to freeze a timestamped, source-cited audit workpaper.
125 total vendors (68 AI tools and 57 enterprise SaaS platforms) scored on 28 public trust criteria with cited evidence trails and side-by-side comparisons.
Plain-language field guides written for security, legal, and procurement teams—covering SOC 2 Type II, HIPAA BAAs, Canadian residency (PIPEDA/Law 25), and EU AI Act audits.
Browse all buyer guides →What each attestation or statute actually demands, who audits or enforces it, what artifact you should request, and which vendors in the Index hold it.
scheduling · verified 2026-10-18
collaborative design · verified 2026-10-18
ai · verified 2026-10-18
collaboration · verified 2026-10-18
observability and security · verified 2026-10-18
customer messaging · verified 2026-10-18
For small and mid-market teams based in—or selling into—the US, Canada, and the EU. Walk through a live vendor tier check, review a frozen audit workpaper for your stack, or scope how our free SOC 2 Compliance Kits fit your upcoming audit window.
[PRIVACY] Zero tracking pixels on page load. Clicking 'Load live calendar' connects directly to our scheduling provider (Calendly) in an isolated frame.
Ready when you are
We block third-party calendar scripts on initial load so your visit stays private and fast. Click below to initialize the interactive booking calendar.
Vendors cannot pay to alter a score, remove an incident, or gate a template. Read the public methodology behind our grades, evidence multipliers, and workpaper verification rules. Not legal or compliance advice.