Core UK Pillar Guide7 min read
Instead of one horizontal AI statute, the UK enforces AI compliance across five enforceable legal pillars: the Data (Use and Access) Act 2025 (substantive Articles 22A–22D in force from 5 February 2026 via SI 2026/82), ICO UK GDPR & Agentic AI expectations, PRA SS1/23 & FCA Consumer Duty (PRIN 2A) in financial services, MHRA AI Airlock Phase 3 & NHS DCB0129 in healthcare, and Equality Act 2010 Sections 19, 20 & 60 in employment.
Read Full UK Guide →ICO & UK GDPR Checklist6 min read
The ICO enforces mandatory AI Data Protection Impact Assessments under UK GDPR Article 35(3)(a) (paired with formal Legitimate Interests Assessments where relied on), strict DUAA 2025 Article 22C contestability workflows, an internal complaints channel (since June 2026), and—following its October 2026 Agentic AI Call for Evidence—named human owners plus PECR Regulation 6 terminal consent wherever AI agents store local session state.
Read Full UK Guide →Cross-Border Compliance6 min read
Yes. Brexit provides zero exemption when UK companies sell AI software into the EU or generate AI outputs used in the EU (up to €35m or 7% global turnover penalties). Under the EU Digital Omnibus (Regulation (EU) 2026/1744), standalone Annex III high-risk deadlines moved to 2 December 2027 and Annex I embedded systems to August 2028—but Article 4 AI literacy and Article 50 GenAI/deepfake transparency obligations took effect on 2 August 2026.
Read Full UK Guide →ISO/IEC 420017 min read
As of late 2026, five certification bodies hold active UKAS accreditation for ISO/IEC 42001 under mandatory ISO/IEC 42006:2025 competence rules: BSI (0003), NQA (0015), ISOQAR (0026), LRQA (0001), and Intertek (0014). Certificates from bodies lacking ISO/IEC 42001 on their UKAS schedule—or relying on foreign accreditations such as DNV under Dutch RvA C024—risk summary rejection on UKAS CertCheck under the Procurement Act 2023.
Read Full UK Guide →Template & Shadow AI6 min read
To pass UKAS ISO/IEC 42001 Stage 2 audits (Clauses 4.3 & 8.1) and satisfy ICO October 2026 Agentic AI expectations, a UK AI Acceptable Use Policy must go beyond generic "don’t paste PII" rules: assign named owners to AI agents, respect PECR Regulation 6 terminal consent, prohibit Equality Act Section 60 pre-offer recruitment questions, enforce DUAA 2025 Articles 22A–22C contestability, and pair policy sign-off with runtime browser DLP such as Harmonic Security.
Read Full UK Guide →2026 Buyer Guide6 min read
UK G-Cloud 14/15 and cloud marketplace filings expose real 2026 pricing tiers: UK SMEs pay £1,495–£1,995/yr for Naq or £3,000–£5,000/yr for ISMS.online; Drata starts at £6,300/yr on G-Cloud (vs Vanta ~$10k + add-on fees); Saidot publishes €1,400–€3,100/mo tiers; Harmonic Security lists $163/user/yr ($32,600 floor) on AWS Marketplace; Aveni Advisory modules run £99–£299/user/mo while Aveni Detect, Warden AI, Mindgard, Hardian Health, Credo AI & Holistic AI remain NDA quote-only.
Read Full UK Guide →2026 UK Pricing Guide5 min read
Direct UKAS certification body audit fees in 2026 depend strictly on ISO/IEC 42006:2025 Table A.1 Effective AI Headcount and Producer/User role—typically 3.5–5.5 days (10 AI staff), 7–11.5 days (50 AI staff), or 16–28 days (1,000 AI staff) billed at £1,050–£1,350/day (NQA, ISOQAR) or £1,350–£1,500/day (BSI, LRQA). A 50-staff AI Provider pays ~£9,600 for initial Stage 1+2 audits, or ~£20,000+ across the full 3-year surveillance cycle before software.
Read Full UK Guide →Public Sector Tenders7 min read
Winning UK central government, NHS, or council AI contracts in late 2026 requires pre-clearing four statutory gates: Central Digital Platform (Find a Tender) registration under the Procurement Act 2023, Cyber Essentials Plus (PPN 014) + PPN 017 AI bid disclosures via G-Cloud 15 (RM1557.15) or AI DPS (RM6200), DSIT ATRS v2.1 algorithmic documentation, and—for health buyers—DTAC v2.0, DCB0129 Clinical Safety Officer sign-off, and CAF-aligned DSPT v8.
Read Full UK Guide →Engineering & AppSec7 min read
Satisfying ICO Agentic AI guidance (October 2026), NCSC Guidelines for Secure AI System Development, and EU AI Act Article 15 requires a three-layer engineering stack: (1) CI/CD red-teaming via Promptfoo (MIT), Giskard (Apache 2.0), or Mindgard generating SARIF 2.1.0 outputs, (2) sovereign UK/EU runtime guardrails via AWS Bedrock Guardrails (eu-west-2 London, $0.10–$0.15/1k units), Azure Prompt Shields (uksouth), or Lakera Guard (<50ms, 92.5% PINT), and (3) automated telemetry ingestion into Drata, Vanta, OneTrust, or Holistic AI.
Read Full UK Guide →