United Kingdom

UK AI Cyber Security Code of Practice

53 controls. 0 other frameworks in our corpus share controls with it. Here is all of it, and how much of it you are already doing.

Page built . This page is derived from the framework corpus, which changes when the corpus is extended rather than daily.

53 controls 0 frameworks share controls with it United Kingdom verified against its source document

Every control below is one this framework asks for. The right hand column counts how many other frameworks in our corpus carry the same control, which is the difference between doing this work once and doing it again for the next standard.

There is no implementation kit for this framework yet. The control list and the overlap above are free and complete.

What you already have

Frameworks whose controls overlap this one, most first. If you run any of them, the count is roughly what you have already evidenced.

No measured overlap with another framework in the corpus.

Every control

CodeControlAlso in
1.1AI security content in the cyber security training programme0
1.2Ongoing staff awareness of AI threats, with secure AI coding training for developers0
10.1Tell end-users where and how their data is used, accessed and stored0
10.2Accessible guidance for end-users, including limitations and security updates0
10.3Support end-users and affected entities through incidents0
11.1Provide security updates and patches, with contingency where updates are impossible0
11.2Treat major updates as a new version and support operators through model changes (11.2 and 11.3)0
12.1Log system and user actions0
12.2Analyse logs for anomalies, breaches and unexpected behaviour0
12.3Monitor internal states of the AI system where useful for security0
12.4Monitor model and system performance over time for security-relevant change0
13.1Secure disposal when training data or a model is transferred0
13.2Secure deletion when a model or system is decommissioned0
2.1Assess the need, the risks and the mitigations before building an AI system0
2.2Design for resilience against adversarial attack, unexpected input and failure0
2.3Create an audit trail for the AI system0
2.4AI-specific risk assessment and due diligence for external components0
2.5Match intended use to data sensitivity, and encourage staff to report AI security risks0
2.6Least-privilege, risk-assessed permissions for the AI system on other systems0
2.7Due diligence on external providers0
3.1Threat modelling and risk management for AI systems0
3.2Pass unresolved AI threats down the chain to operators and end-users0
3.3Assurance over external parties responsible for AI risks0
3.4Continuous monitoring and review of system infrastructure against risk appetite0
4.1Build in and maintain capabilities for human oversight0
4.2Make outputs easy for responsible humans to assess0
4.3Technical measures where human oversight is a risk control0
4.4Verify the data custodian's specified controls are implemented0
4.5Tell end-users which uses of the system are prohibited0
5.1Comprehensive asset inventory including interdependencies0
5.2Processes and tools to track, authenticate, version and secure AI assets0
5.3Disaster recovery plans tailored to attacks on AI systems0
5.4Protect sensitive data and sanitise inputs, with proportionate protection of weights0
6.1Evaluate access control frameworks for APIs, models, data and pipelines0
6.2Controls on externally offered APIs against attacks on the AI system0
6.3Dedicated, segregated environments for development and model tuning0
6.4Publish a clear vulnerability disclosure policy0
6.5AI system incident management plan and recovery plan0
6.6Cloud service contracts that support these requirements0
7.1Secure software supply chain processes for AI development0
7.2Justify, mitigate and disclose use of poorly documented or unsecured components0
7.3Re-run evaluations on released models before use0
7.4Tell end-users before models are updated0
8.1Document system design and maintenance plans, and publish component hashes0
8.2Record the provenance and use of public training data0
8.3Audit log of changes to system prompts and model configuration0
9.1Security assessment testing before release of models, applications and systems0
9.2Operator testing before deployment, using independent skilled testers0
9.3Share testing and evaluation findings with system operators0
9.4Evaluate outputs for leakage of model internals and unintended influence0
GThe Implementation Guide: worked scenarios and example measures0
RStakeholder roles0
SScope and status of the Code of Practice0

Tell me when UK AI Cyber Security Code of Practice files something new

One email when a public company newly discloses something this framework governs, naming the company and what our corpus says it puts in scope. Nothing else, and one click to stop.

What this page is

A control-level reference for UK AI Cyber Security Code of Practice, drawn from our framework corpus. Control codes and titles are references to the standard, not reproductions of it. The overlap counts and the auditor artefacts are our own work and are the part you will not find elsewhere.

Measure this against what you already run · All frameworks · Today's edition