keepcalculator

06 Numbers and Powers

UK Cyber Security Bill vs EU AI Act: Which Rules Apply?

Card No. 06·77733 · Updated · Standard formula, shown below

Which rules apply depends on what your business does, where it operates and whether it provides or uses regulated technology: the UK Cyber Security and Resilience Bill concerns cyber-security duties for specified services, while the EU AI Act regulates AI systems and their providers, deployers and other participants. A UK business can be subject to the EU AI Act when it places an AI system on the EU market or uses one whose output is used in the EU; the Bill’s reach is tied to UK law and the services it covers.

Two different laws with different purposes

The comparison is not a choice between two versions of the same rulebook. The UK bill is intended to strengthen the cyber resilience of organisations and services considered important to the UK economy and society. It builds on the existing UK framework for network and information systems, commonly called the NIS Regulations, including oversight of operators of essential services and relevant digital service providers.

The EU AI Act is a product- and use-focused law. It establishes requirements for AI systems according to their risk and assigns duties to actors such as providers, deployers, importers and distributors. It can apply beyond the EU: a company established in the UK may be in scope if it places an AI system on the EU market, puts it into service there, or its AI system’s output is used in the EU.

Legal status matters. A bill is proposed legislation; it does not create enforceable duties merely because it has been introduced or discussed. Check the bill’s current progress, final enacted text and commencement provisions before treating its proposed requirements as obligations. The EU AI Act, by contrast, entered into force in 2024, but its provisions apply in stages. The applicable date and requirements depend on the provision and the system or activity involved.

What the UK Cyber Security and Resilience Bill is about

The UK government’s bill is designed to update and expand the UK’s cyber-security and incident-reporting regime. Its direction includes bringing more relevant organisations and services within regulatory oversight, strengthening regulators’ powers and improving visibility of significant incidents and supply-chain risks. The precise scope and obligations depend on the final legislation and implementing rules, so businesses should not assume that every UK company will be covered—or that proposals are already law.

For a business, the practical first question is whether it operates a service or supplies a relevant service that falls within the regime, and which regulator would oversee it. The existing NIS framework covers defined categories rather than all businesses generally. Proposed changes may alter coverage, but a company should map its activities, customers, infrastructure and supply-chain role against the enacted rules when available.

Even if a business is not directly regulated, it may face indirect pressure. A regulated customer could ask suppliers for security controls, incident notification, audit information or contractual assurances. Those requests arise through commercial and supply-chain relationships; they do not by themselves make the supplier a regulated entity under the bill.

Who and what the EU AI Act regulates

The AI Act applies to AI systems, not to every software product simply because it uses automation. It sets different obligations depending on an actor’s role and the system’s risk category. A provider develops an AI system or has it developed and places it on the market or puts it into service under its own name. A deployer uses an AI system under its authority, subject to the Act’s definitions and exceptions.

  • Prohibited practices: Certain AI uses are banned, subject to the exact statutory provisions and exceptions.
  • High-risk systems: Systems in specified areas or categories face extensive requirements, which can include risk management, data governance, documentation, logging, human oversight, accuracy, robustness and cybersecurity.
  • Transparency duties: Particular systems, such as certain chatbots or systems that generate or manipulate content, may have disclosure duties.
  • General-purpose AI: The Act sets obligations for providers of general-purpose AI models, with additional requirements for models presenting systemic risk.

Using an AI tool does not automatically make a company its provider, and not every AI system is high-risk. But a deployer can have duties, especially when using a high-risk system in a covered setting. A business that substantially modifies a system or changes its intended purpose may also need to examine whether its legal role changes.

How to decide which rules apply

Assess the two regimes separately, then coordinate the results. A practical initial review can follow these steps:

  1. Map your activities and locations. Identify where your services are supplied, where customers and users are located, and where AI outputs are used.
  2. Check UK cyber-regime coverage. Compare your service and organisational role with the current NIS rules and the final status, scope and commencement of the UK bill.
  3. Inventory AI systems. Record each system’s purpose, provider, users, affected people, deployment locations and whether it is bought, adapted or developed in-house.
  4. Classify your AI role and risk. Work out whether you are a provider, deployer, importer or distributor, and whether the use falls within a prohibited, high-risk, transparency or general-purpose AI provision.
  5. Translate requirements into controls. Assign owners for security, procurement, records, incident handling, human oversight and supplier coordination, with dates based on the rules that actually apply.

For example, a UK cloud company serving UK clients should examine whether its service falls within UK cyber-security legislation. If it also supplies an AI screening system to an EU employer, the EU AI Act may apply to the AI activity, depending on the system’s purpose and the company’s role. The cloud service and the AI screening tool may therefore trigger different analyses, even inside one business.

Overlap, differences and common misconceptions

Cybersecurity is relevant to both, but they do not impose identical obligations. The UK bill centres on resilience and incident reporting for covered services. Under the AI Act, cybersecurity is one part of the safety and robustness requirements for certain AI systems, particularly high-risk systems; it is not a general cyber-security law for every organisation that uses AI.

One common misconception is that compliance with one regime automatically satisfies the other. Good security practices—access control, vulnerability management, supplier checks, incident procedures and reliable records—can support both, but each law has its own scope, definitions, evidence and reporting channels. Another misconception is that being based outside the EU rules out the AI Act. Market placement, use and the location of output can matter more than a company’s headquarters.

There can also be separate rules beyond these two. UK data protection law, EU data protection law, sector-specific regulation and contractual commitments may apply to the same product or incident. A cyber incident involving personal data, for example, can raise distinct data-protection notification questions. Do not treat a single compliance checklist as a substitute for checking each applicable regime.

Frequently Asked Questions

Does the UK Cyber Security and Resilience Bill apply to every UK business?
No. It is aimed at specified services and organisations, with scope determined by the legislation and related rules. Check the bill’s current legislative status and final coverage rather than assuming all businesses are included.

Does the EU AI Act apply to a UK company?
It can. A UK company may be in scope if it places an AI system on the EU market, puts it into service in the EU, or the system’s output is used there, depending on the facts and the Act’s provisions.

Does using ChatGPT or another AI tool make my business a provider?
Not necessarily. A business that uses a third-party system is often acting as a deployer, but its role depends on what it does, how the system is supplied and whether it modifies or rebrands it. Check the statutory definitions and the actual arrangement.

Are the UK bill’s duties already enforceable?
A proposal is not enforceable simply because it is called a bill. Confirm whether it has become law, when relevant provisions commence, and whether regulations or guidance affect the duties.

Does the AI Act replace cybersecurity or data-protection law?
No. It adds AI-specific requirements. Cyber-security and data-protection laws can apply alongside it, as can sector rules and contractual requirements.

What businesses should do now

Keep a current register of services, suppliers and AI systems, and note the business purpose and countries involved. Follow the UK bill through enactment and commencement rather than acting on proposals as though they were final law. For EU-facing AI, identify the system’s role and risk category early, because provider and deployer responsibilities differ and some AI Act requirements are phased. Where a system could be high-risk, or the business model makes its legal role uncertain, obtain advice based on the final text and specific facts.

Related

Method: the standard formula for this calculation, rounded as each result states. See how the calculators work or report an error.

More cards in this drawer

The whole Numbers and Powers drawer →