Ensure AI systems are safe, secure and robust throughout their lifecycle

Practical action

Run lifecycle risk assessments and keep security testing evidence that you can show a regulator.

Who does it apply to?

Non-binding expectation for all AI developers and deployers in regulated activities.

Actors

Provider / developerDeployer / user organisation

Sectors

Cross-sector / all sectors

Use cases

Generative AI and foundation modelsSafety-critical and infrastructure

Which controls meet this duty?

Satisfies: the control, operated properly, does the work the duty asks for. Supports: it contributes but the duty needs more. Each control page lists every other duty it serves, so work done once can be counted once.

• satisfiesProcessAI system owner · once per ai system
  AI risk assessment and lifecycle risk register

  Serves 12 recorded duties · evidence: AI system risk assessment, Per-system AI risk register, Residual-risk acceptance

  Lifecycle identification and management of risks.

• satisfiesTechnical measureQuality or testing lead · at launch and on material change
  Accuracy, robustness, fairness and security testing

  Serves 15 recorded duties · evidence: Pre-release test report, Test plan and acceptance criteria, Release test sign-off

  Security and robustness evidence to show a regulator.

Templates that cover this duty

Generated from the records. Free · no account · links emailed to your work address. This duty is cited in each file with its source reference and a link back here.

• AI Risk Register (MIT AI Risk Repository taxonomy)

  Register · XLSX · v4

• AI Impact Assessment

  Assessment · DOCX and XLSX · v5

• ISO/IEC 42001 Gap Assessment and Statement of Applicability

  Assessment · XLSX · v3

• NIST AI RMF ↔ EU AI Act ↔ ISO/IEC 42001 Crosswalk

  Crosswalk · XLSX · v3

All 6 templates that cover it →

What evidence would a reviewer expect?

Framework mappings

Original editorial crosswalks. They cite clause numbers only and reproduce no standard text; confidence reflects how direct the mapping is.

See every United Kingdom duty mapped this way →

Frequently asked questions

Cite this record

AIPolicyTracker (2026). “Ensure AI systems are safe, secure and robust throughout their lifecycle (UK AI regulation framework)” [AIPT-OBL-0072, verified 26 September 2026 by Bhaskar Bhatt]. https://aipolicytracker.org/id/AIPT-OBL-0072 (accessed 27 October 2026). Data licensed CC BY 4.0.

Permanent id AIPT-OBL-0072: it never changes or gets reused, and its address keeps working if this page moves.

Cite the official text alongside it: A pro-innovation approach to AI regulation, UK Government, Department for Science, Innovation and Technology, https://www.gov.uk/government/publications/ai-regulation-a-pro-innovation-approach.

@misc{aipolicytracker2026_ensureaisystemsaresafese,
  author       = {{AIPolicyTracker}},
  title        = {{Ensure AI systems are safe, secure and robust throughout their lifecycle (UK AI regulation framework)}},
  year         = {2026},
  howpublished = {\url{https://aipolicytracker.org/obligations/uk-principles-safety-security-robustness}},
  url          = {https://aipolicytracker.org/obligations/uk-principles-safety-security-robustness},
  urldate      = {2026-10-27},
  note         = {Data licensed CC BY 4.0. Accessed 27 October 2026}
}

Change history

Similar obligations in other instruments

• Deployers must implement a risk management policy and programme — Colorado AI Act, Colorado (United States)
• Apply minimum risk-management practices to high-impact AI — OMB M-25-21, United States
• Operators of high-impact AI must establish and operate a risk management plan — AI Basic Act, South Korea
• Establish a risk management system for high-risk AI — EU AI Act, European Union
• Developers must use reasonable care to avoid algorithmic discrimination — Colorado AI Act, Colorado (United States)
• Deployers must use reasonable care to avoid algorithmic discrimination — Colorado AI Act, Colorado (United States)
• Prioritise, respond to and monitor AI risks (Manage) — NIST AI RMF, United States (voluntary)