Stay updated with the latest industry insights on AI compliance.

The International AI Safety Report 2026, read as compliance evidence: what it says, where it fits under Article 55 and the GPAI Code, and what to log.

AI liability turns on evidence from 9 December 2026: what the new EU Product Liability Directive presumes, who answers, and the file to keep ready.

ISO/IEC 23894 explained: structure, annexes, how its ISO 31000 definition of risk diverges from EU AI Act Article 9, prEN 18228 and a five-step evidence method.

ISO/IEC 22989 defines 117 AI terms and is free, but provider, autonomy and risk mean something else in the EU AI Act. What is inside and how to use it.

The OWASP Top 10 for LLM changed in August 2026. See the new order, what moved, and how each entry maps to the EU AI Act, ISO/IEC 42001 and audit evidence.

An AI BOM lists the models, datasets and dependencies of an AI system. See the 50 G7 minimum elements and what the EU AI Act and CRA add on top.

What a model card contains, what the Hugging Face template misses against EU AI Act Annex IV and XI, and six steps to turn a card into audit evidence.

The OECD AI Principles are not law, yet they define what an AI system is under the EU AI Act. What they say, what changed in 2026, and how to evidence them.

GDPR-compliant AI in 2026: legal basis after EDPB Opinion 28/2024, web scraping guidelines, Article 4a bias data, DPIA vs FRIA, and the 8 records to keep.

Agentic AI security read as an auditor would: the ten OWASP agentic risks mapped to EU AI Act duties, the records that prove control, and the deadlines.

NIST AI 600-1 explained: the 12 generative AI risks, 211 suggested actions, its 2026 status, the Texas TRAIGA defense and the EU AI Act mapping.

Model drift quietly turns a validated AI model into a non-compliant one. Learn the types, detection metrics and what EU AI Act Articles 15 and 72 require.

HIPAA compliance software was built for systems that store PHI, not for systems that infer from it. What a 2026 tool must cover, and what to demand.

An AI inventory is the artefact every AI rule assumes. See which clauses compel one (EU AI Act, NIST, ISO 42001, OMB) and the fields each expects.

China AI regulation explained for foreign firms: CAC filings, AI content labels, companion AI rules, 2026 enforcement and the evidence to keep ready.

California SB 243 set the first companion chatbot rules. Adam's Law (SB 1119) added risk assessments, parental controls and independent audits from 2027.

The California AI Transparency Act took effect on 2 August 2026. What SB 942 requires, what SB 1000 would change, and the records you must keep.

Agentic AI vs generative AI is not just a technical difference. See what changes in classification, oversight, risk and evidence when AI acts.

The CCPA regulations in force since January 2026 add ADMT duties from 1 January 2027, risk assessment filings in 2028 and audits to 2030. Dates and evidence.

Adversarial AI attacks ML models through poisoning, evasion and prompt injection. See the attack types, defenses, and governance controls the EU AI Act now requires.

California AI laws explained by role and date: SB 53, SB 942, SB 243, CCPA ADMT, FEHA rules and the bills Newsom signed in September 2026.

TRAIGA has been in force since January 2026. What the Texas AI law prohibits, how the NIST AI RMF safe harbour works, and the evidence you need to rely on it.

Standard vendor due diligence was built for a pre-AI supply chain. Here are the 12 AI-specific questions to add, and the legal duty behind them.

Model risk management is being rewritten for AI. See how SR 26-2, the EU AI Act, ISO 42001 and NIST AI RMF reshape MRM for machine learning and GenAI.

Policy management software must now prove AI policies work, not just that staff signed them. Evaluation criteria, EU AI Act duties and buying traps.

Human oversight is an EU AI Act Article 14 obligation, not a principle. What providers must build, what deployers must staff, and when it applies.

An AI governance committee only counts if it survives an audit. The clauses behind it, the seats, the decision rights and the records you must keep.

A governance risk and compliance tool must now inventory AI systems, map EU AI Act duties and hold audit-ready evidence. Here is the capability checklist.

HIPAA compliance AI goes far beyond a BAA. See how Section 1557, HTI-1, FDA and the EU AI Act govern clinical AI, and what evidence auditors expect.

The EU AI Act high-risk deadline moved to December 2027, the classification duty did not. Work through the Article 6 determination and the exemption trap.

EU AI Act Article 50 has applied since 2 August 2026. What providers and deployers must disclose, mark and label, plus the 2 December 2026 deadline.

A data governance framework that satisfies auditors, not just committees: the four pillars re-scoped to EU AI Act Article 10, ISO 42001 A.7 and NIST AI RMF.

Conformity assessment is how a high-risk AI system proves it complies. The Article 43 routes, the Annex IV evidence, and the December 2027 deadline.