You're offline - Playing from downloaded podcasts
Back to All Episodes
Podcast Episode

EU Drops Key AI Act Guidance on High-Risk Systems Ahead of August Deadline

February 4, 2026

Audio archived. Episodes older than 60 days are removed to save server storage. Story details remain below.

The European Commission has published guidelines clarifying how to classify high-risk AI systems under the EU AI Act, with full enforcement of strict compliance requirements set for August 2026. Companies face potential fines of up to thirty-five million euros or seven percent of global turnover for violations.

New Guidelines Define What Counts as High-Risk AI

The European Commission has met a critical February 2026 deadline by publishing guidance on how companies should classify their AI systems under the bloc's landmark Artificial Intelligence Act. The guidelines clarify which AI tools fall into the high-risk category and outline monitoring obligations for providers, marking a pivotal step toward full enforcement in August.

High-risk AI systems are those deployed in sensitive domains where algorithmic outputs can materially affect people's rights or opportunities. This includes AI used in employment decisions, education, access to essential services and benefits, biometrics, critical infrastructure, and certain law enforcement contexts.

Compliance Requirements Are Substantial

Once the high-risk rules take effect in August 2026, providers must meet extensive pre-market obligations. These include conducting risk assessments, ensuring high-quality training datasets to reduce discriminatory outcomes, maintaining detailed technical documentation, implementing human oversight measures, and meeting strict accuracy, robustness, and cybersecurity standards. Systems must also undergo conformity assessments and be registered in an EU database before being placed on the market.

The financial stakes are significant. Penalties for non-compliance can reach thirty-five million euros or seven percent of global annual turnover for banned practices, with lower tiers for other violations. For startups and small businesses, compliance costs for high-risk categories are estimated between one hundred and sixty thousand and three hundred and thirty thousand euros in auditing and legal fees alone.

Digital Omnibus Could Shift Timelines

Adding complexity to the landscape, the Commission's Digital Omnibus proposal from November 2025 could defer some high-risk obligations until supporting standards and guidelines are fully in place, potentially pushing deadlines to late 2027 or even 2028 for certain categories. The proposal aims to reduce administrative burdens by twenty-five percent for all businesses and thirty-five percent for small and medium enterprises.

However, the Omnibus remains a proposal under legislative review, and experts warn companies against delaying compliance preparations. Finland has already activated national supervision laws, becoming the first EU member state with fully operational AI Act enforcement powers.

Published February 4, 2026 at 9:26am

More Recent Episodes