Systems are beginning to harness artificial intelligence and automation to support decision-making at all levels. There is a clear strategy and ethical framework to ensure these tools are used responsibly, augmenting professional judgement and not replacing it. By analysing large volumes of data across the system, AI can help spot issues earlier, predict future needs, and personalise support. The aim is to provide more proactive, preventative care.
Key actions:
- Develop a collaborative AI strategy - Co-produce a plan with all partners for introducing AI and automation, focused on shared priorities where data-driven insight can make a difference (e.g. predicting falls or hospital readmission). Ensure alignment with the system’s vision of better outcomes and include clear goals.
- Establish ethics and governance up front - Set up robust governance (e.g. AI ethics board or forum) alongside agreed data-sharing protocols to ensure AI use is transparent, fair and free of bias. Include individual and patient voices in developing an ethics policy. Define that professionals remain accountable for decisions, with AI providing support.
- Start with high-impact use cases - Launch pilot AI projects focused on pressing challenges such as using machine-learning models on combined data sets to predict individuals at risk (e.g. falls, mental health crisis, or homelessness) for early intervention. Use AI-driven analysis of case notes (natural language processing) to refine the view of challenges facing citizens. Prioritise projects that can show quick wins and tangible benefits.
- Leverage automation for efficiency - Deploy automation tools for routine analysis and triage, freeing up staff time. Implement systems to summarize multi-agency information for a “single view” of a person and use AI rules to flag low-risk referrals for self-help, allowing skilled staff to focus on complex cases. Improve responsiveness and consistency in decision-making.
- Build workforce confidence and skills -Invest in training and engagement so that staff understand and feel comfortable using AI tools, using workshops to show how a predictive dashboard or an AI alert works in practice, emphasising that these tools assist their expertise (e.g. alerting a social worker to a pattern they might not have seen). Encourage feedback from frontline teams to refine the tools.
- Iterate and scale what works - Evaluate the pilots rigorously, based on desired outcomes, and share results openly. Refine models to improve accuracy and fairness. When a pilot proves successful, mainstream it across the system, and keep exploring new AI opportunities (e.g. annual innovation review) to remain up to date on emerging technology, always with clear evidence of benefit.
Ambition level actions
Foundational
- System partners acknowledge AI's potential and are united in principle.
- System understands the need for an initial AI/automation strategy, supported by leadership buy-in and basic data sharing arrangements.
- Efforts are focused on building foundations such as governance, common data infrastructure, and ensuring everyone is on board.
- Staff awareness of AI is beginning, but they may not be clear on exactly how it can improve current working practices.
Developing
- The system has an initial AI plan, which has been agreed with all staff.
- The system has initiated a pilot project using AI to support decisions.
- The project has governance oversight, such as an ethics panel, reviewing the impact and checking for bias.
- Early results are promising, revealing insight to support outcomes.
- Staff involved in the pilot are engaged, receiving training and starting to see how AI-driven insights could be used in daily decisions.
- Staff have awareness of AI and how it can support decision making in theory, but need to see the results in practice.
Established
- A clear, system-wide AI strategy is in action, with some AI tools used across different parts of the system.
- Predictive models are live, and these tools are integrated into regular workflows, with recommendations for front line staff.
- Real examples of impact are evident, and staff at all levels can articulate the value of AI, with managers and frontline workers sharing how examples have improved resource planning and individual care.
- The system continues to monitor ethics and accuracy, checking models for bias and updating them as needed.
Exemplar
- The system is a national leader in ethical AI adoption.
- AI is used by everyone from strategists to support workers as a natural part of their toolkit, always with professional oversight.
- The system leverages technology for citizen-led and community-led prevention, offering apps or wearables to residents and linking community volunteers through an AI-coordinated network.
- Outcomes have visibly improved in areas where AI-supported interventions are used, with fewer people in crisis, more early help, and better experiences for service users.