VerifyWise's AI Governance Library

VerifyWise, a vendor of an AI governance software platform, maintains a public directory of AI governance reference material — regulations, standards, frameworks, and tools — described on the site as covering 543 resources across 24 categories at the time of retrieval (2026-08-17). A separate, smaller companion directory on the same site focuses specifically on agentic AI (roughly 121 resources across 8 categories).

Category structure

The library organises resources into categories including: regulations and laws; standards and certifications; governance frameworks; internal policy and governance; risk taxonomies; assessment and evaluation methods; transparency and documentation; incident and accountability practices; organisational roles and processes; sector-specific governance; international initiatives; ethics and principles; tooling and implementation; academic and research references; datasets and benchmarks; and open-source governance projects.

Examples of listed resources

The featured and recently added entries captured at retrieval time included:

• EU AI Act (official text) — described as the first comprehensive AI-specific legal framework, using a risk-tiered approach (prohibited, high-risk, limited-risk, minimal-risk) with obligations for high-risk systems around risk management, data governance, transparency, and human oversight.
• OECD AI Principles — a explainer of the five principles agreed in 2019 and adopted by 46-plus countries, with notes on how they relate to the EU AI Act and ISO/IEC 42001.
• California SB 243 (Companion AI Guardrails Act) — signed into law in October 2025 and effective January 2026, it requires operators of companion-style AI chatbots to disclose that a user is interacting with AI, restrict sexually explicit content for minors, provide crisis resources for self-harm-related conversations, and report annually to a state suicide-prevention office; it also creates a private right of action for people harmed by non-compliance.
• OpenAI's agentic-AI safety practices — a set of seven practices for operating autonomous AI agents safely (assigning accountability, logging actions, requiring human approval at key points, bounding what an agent can do, rolling out capabilities in stages, designing for reversibility, and building in shutdown controls).
• OWASP AI Bill of Materials (AIBOM) — a proposed standard format for documenting an AI system's components, training-data sources, model provenance, and security configuration, modelled on how software bills of materials (SBOMs) standardised supply-chain transparency for conventional software.
• C2PA Content Credentials — a technical specification for cryptographically attaching provenance metadata to digital media, so that a file's origin and edit history (including AI generation) can be verified; backed by a coalition of large technology and media companies.
• AIGP certification (IAPP) — a practitioner certification covering AI governance, with a cited exam cost in the $550–$750 range, 100 questions across four knowledge domains, and a stated pass mark.

What was left out of this summary

VerifyWise's own product marketing — its platform feature list, "get a demo" prompts, comparison pages against competing vendors, and links to its blog, social channels, and policy templates — has been excluded as vendor self-promotion unrelated to the factual content of the directory. Resource descriptions above have been paraphrased in original wording rather than copied from the source; where a specific figure (dollar amount, percentage, date) is cited, it reflects what the source page stated and should be independently verified before being relied upon, since this file is a secondary summary, not the primary source.