Regulatory guidance, technical analysis and AI governance strategy for regulated organisations.
Start here: EU AI Act Compliance: a guide for UK organisations. Who it applies to, the deadlines, the penalties, and how to prepare.
Grace Blackwell will be on stand A10 at the NPCC Innovation and Digital Summit in Liverpool from 20 to 22 October 2026, showing TOPAS for Word and offering a 14-day trial.
Read article →The EU AI Act's high-risk deadlines moved to 2027 and 2028. Several duties did not move at all. A short checklist of what still applies now.
Read article →After a May agreement, the EU AI Act simplification package has been formally adopted. The high-risk deadlines have moved. Here are the dates that changed.
Read article →The UK's Digital Regulation Cooperation Forum has been asking how people view the risks of generative and agentic AI. The questions signal where scrutiny is heading.
Read article →The EU AI Act, ISO 42001 and the NIST AI RMF overlap far more than they differ. Mapping them once removes most of the duplicated compliance work.
Read article →European providers are launching sovereign cloud and AI services built for EU data rules. For UK regulated buyers, the test is proof, not the label.
Read article →From 2 August 2026 the European Commission can enforce the AI Act's rules on general-purpose models, with fines up to 3 percent of global turnover. What providers face.
Read article →The EU AI Act's toughest duties fall on the high-risk systems defined in Annex III. A recent delay buys time to prepare. How to tell whether yours qualifies.
Read article →The UK is testing AI rules through regulatory sandboxes rather than a single AI Act. AI Growth Labs start with legal services. What it means for regulated firms.
Read article →The EU's proposed Cloud and AI Development Act aims to speed sovereign compute. UK firms sit outside the framework, but the governance signal still applies.
Read article →SpaceX, OpenAI and Anthropic are nearing some of the largest listings ever. The valuations rest on one expectation about AI, with real concerns underneath.
Read article →Anthropic, SoftBank and a stalling EU programme in one fortnight. AI capital is concentrating, and for regulated buyers that is a governance question.
Read article →The EU's €20 billion plan for five AI gigafactories is slipping as bidder interest falls. European public compute is further off than the headlines suggest.
Read article →SoftBank will invest up to €75 billion in AI data centres in France, dwarfing the EU's own programme. Where Europe's compute will sit, and why it matters.
Read article →Anthropic has raised $65 billion at a $965 billion valuation, passing OpenAI. What the round signals for a concentrating market, with a note from us.
Read article →Data that mimics the shape of real records without containing real people. It keeps personal data out of testing, when generated with care.
Read article →Where a model runs decides which jurisdiction governs the data and how quickly it responds. A governance decision as much as a technical one.
Read article →The first international standard for governing AI across its lifecycle. It complements EU AI Act preparation rather than replacing it.
Read article →Upload your AI project files and get an instant compliance pre-check. Supports Python, TypeScript, and .NET across 10 frameworks. No account required.
Read article →The EU has deferred high-risk AI deadlines by up to 16 months. Standalone Annex III obligations move from August 2026 to December 2027.
Read article →Compliance questionnaires produce confident wrong answers for agentic AI systems. Technical scanning catches architecture-level risks that no checklist can surface.
Read article →DORA has been in force since January 2025. AI systems within financial entities create specific ICT risk management obligations most firms haven't addressed.
Read article →Fines up to €35 million or 7% of global turnover. Understanding which violations trigger which tier is essential for risk prioritisation.
Read article →Agentic AI systems fall within the EU AI Act's scope. Classification depends on domain and decision type, not the orchestration framework.
Read article →The EU AI Act has extraterritorial reach that catches UK companies through three provisions in Article 2(1). The UK has no equivalent horizontal legislation.
Read article →