Explore our new Cybersecurity Terms Glossary, explained by our experts.

View the Glossary View the Glossary
PRIVIA

Cybersecurity Terms Glossary

What Is the EU AI Act?

The EU AI Act is the first comprehensive AI law, classifying systems by risk and imposing duties on providers and deployers, including UK organisations serving the EU market.

Illustration for the cybersecurity term EU AI Act
Home / Blog / Cybersecurity Terms / What Is the EU AI Act?
17 November 2026

What Is the EU AI Act?

What Is the EU AI Act?

Regulation (EU) 2024/1689, generally called the EU AI Act, entered into force on 1 August 2024 and is the first broad legal framework for artificial intelligence anywhere. It regulates by use rather than by technology: the same model can be unregulated in one application and tightly controlled in another, depending on what decision it influences and who is affected. For UK organisations the relevant point is extraterritorial scope—the Act applies to providers placing AI systems on the EU market and to deployers established in the EU, and also where the output of a system is used in the EU, regardless of where the provider sits.

Four Risk Tiers

The Timetable

DateWhat applies
2 February 2025Prohibited practices, and the AI literacy duty for providers and deployers
2 August 2025Obligations for general-purpose AI models, governance structures, and penalties
2 August 2026The bulk of the Act, including Annex III high-risk duties and transparency marking
2 August 2027High-risk AI embedded in regulated products under Annex I

What High-Risk Status Requires

Providers of high-risk systems must operate a risk management system across the lifecycle, meet data governance and quality requirements, maintain technical documentation, enable automatic logging with records kept for at least six months, provide instructions that allow deployers to use the system correctly, design for human oversight, and achieve appropriate accuracy, robustness, and cybersecurity. Conformity assessment and CE marking follow, with registration in an EU database. Article 15 makes cybersecurity an explicit legal requirement, naming resilience against data poisoning, model poisoning, adversarial examples, and confidentiality attacks—the first time these appear as compliance obligations rather than good practice. Deployers have their own duties: use the system as instructed, assign competent human oversight, monitor operation, keep logs, and in several public-sector cases carry out a fundamental rights impact assessment.

General-Purpose Models and Penalties

Providers of general-purpose models must publish technical documentation, a summary of training content, and a copyright policy. Models presenting systemic risk face additional duties including adversarial testing, incident reporting, and cybersecurity protection for the weights themselves. A Code of Practice agreed in 2025 gives a presumption-of-conformity route. Penalties are tiered: up to €35 million or 7% of worldwide annual turnover for prohibited practices, up to €15 million or 3% for most other breaches, and up to €7.5 million or 1% for supplying incorrect information.

How This Interacts With UK Rules

The UK has no equivalent statute and continues with a sector-regulator approach, so a UK organisation’s binding domestic obligations still come from UK GDPR, the NIS Regulations, FCA and other sector rules, and general consumer and equality law. The Act becomes relevant through markets and supply chains: selling an AI-enabled product into the EU, providing a system an EU client deploys, or supplying a component to a provider who must evidence conformity. Many UK organisations are choosing to align voluntarily, because the Act’s documentation, logging, and testing expectations map closely onto AI auditing and traceability work they need anyway, and because ISO/IEC 42001 provides a management system that satisfies both audiences. At Privia Security, our Regulation and Compliance Services help organisations determine where they sit in the Act and what evidence they will need.

FAQ – Frequently Asked Questions

Does the EU AI Act Apply to UK Companies?

It can. The Act reaches providers placing AI systems on the EU market, deployers established in the EU, and cases where a system’s output is used in the EU even though the provider is elsewhere. A UK business selling AI-enabled software to EU customers, or processing for an EU client, should assume it is in scope and establish which role it holds.

Is Our Chatbot High-Risk?

Usually not. A customer service chatbot generally falls under the transparency tier, requiring users to be told they are speaking to an AI system. It becomes high-risk if it performs a function listed in Annex III, such as assessing creditworthiness, screening job applicants, or determining access to essential services. The function decides the tier, not the technology.

What Does Article 15 Require for Cybersecurity?

That high-risk systems achieve an appropriate level of accuracy, robustness, and cybersecurity, and resist attempts to alter their use or performance. The Act names poisoning of training data and models, adversarial inputs, and confidentiality attacks as threats to be addressed. In practice that means documented threat modelling, adversarial testing, and controls over data and model integrity.

Will the UK Introduce Something Similar?

No comprehensive statute has been enacted. Policy has centred on existing regulators applying existing law, supported by the AI Security Institute and voluntary guidance such as the code of practice issued in January 2025. Targeted legislation on specific harms remains possible, so organisations should track announcements rather than assume the position is settled.

Related Terms

You May Be Interested In These

What Are Disk Imaging and Write Blockers?

Cybersecurity Terms

17 November 2026

What Are Disk Imaging and Write Blockers?

Read More
What Is the Cost of Production Downtime?

Cybersecurity Terms

17 November 2026

What Is the Cost of Production Downtime?

Read More
What Is Malware?

Cybersecurity Terms

17 November 2026

What Is Malware?

Read More