Delivery, assurance and operations

• Do AI projects still need to meet the government Service Standard?

  6 sources

• Do I need to tell users when a service uses AI?

  4 sources Contrasting

• How should I do user research for an AI product?

  1 source

• How do I monitor an AI system once it is live?

  1 source

• Do staff need AI-specific cyber security training?

  3 sources Contrasting

• How should I secure the AI supply chain (models and components)?

  4 sources

• What AI assets should I document and inventory?

  2 sources

• How should I decommission AI models and training data?

  2 sources

• What happens if the model drifts or starts performing worse over time?

  1 source

• How do I keep an inventory of AI systems in my organisation?

  2 sources

• Who is accountable if an AI system causes harm or makes a bad decision?

  4 sources

• How do people challenge or seek redress for an AI-influenced decision?

  3 sources

• Which AI assurance techniques should my organisation use?

  2 sources

• How do I keep AI assurance proportionate to risk?

  1 source

• What governance should underpin AI assurance?

  1 source

• When should we use third-party conformity assessment for AI?

  1 source

• Should we red-team AI before release?

  1 source

• Should we document AI models with model cards or SBOMs?

  1 source

• Should we monitor AI system inputs as well as outputs?

  2 sources

• How do teams evidence Secure by Design?

  1 source

• Who should own cyber security risk for an AI service under Secure by Design?

  1 source

• What is continuous assurance under Secure by Design?

  2 sources

• Do equality and data-protection checks stop once an AI contract is signed?

  1 source