Our AI Strategy for 2025 - 2028

Since our inception in 1852 our staff have been instrumental in using ground-breaking research and the latest technology to revolutionise the care of the children and young people we care for. This, our first, Artificial Intelligence (AI) strategy builds on this legacy and sets out how we aim to harness this technology in an ambitious yet thoughtful and responsible way. It sets out to undertake two broad and interrelated objectives: to enhance patient care and operational efficiency. It is launched at the foothills of what is expected to be a global AI revolution: all around us this technology is being used to enhance our daily lives.

We are children and young person’s hospital and they expect us to use AI to make their care and experience better. But they are also clear that it cannot do everything and the contact they have with our teams, and the sense of community they feel is a vital part of their care journey. This strategy sets out how we will use AI to enhance clinical decision making, particularly in areas such as imaging, early disease detection, personalise treatments and patient monitoring. Integrating AI into our electronic patient record will reduce administrative burden on our staff which we anticipate will free up their time to focus more on direct patient care.

We currently operate in a resource constrained healthcare sector which is having a prolonged impact on staff well-being and morale. We have a duty to consider how we can harness the technology to carry out more repetitive tasks that can be done in the fraction of the time to free up staff time and resources. On an organisational level AI will be able to provide better insight in how we can run our hospital and optimise hospital operations through AI-driven scheduling, resource planning and cost efficiency.

We have not created this strategy alone. It has been developed through extensive dialogue with key stakeholders including staff, children and young people's representatives and non- executive directors. It reflects a broad understanding of AI's potential benefits and acknowledges the concerns around its ethical and responsible delivery. This is such an exciting area but we recognise that AI is not ready to make autonomous decision making and we will harness its potential in a responsible and ethical way. This means ensuring all our initiatives protect patient and staff data, are sustainable from both a financial and environmental perspective and of course meet regulatory requirements.

The potential benefits of AI are so great that we will establish a dedicated Department for Advanced Analytics and AI and we will upskill our workforce on safe and ethical AI use to ensure that our staff are well-equipped to work with these advanced technologies. We will also bring the outside world in by working to foster meaningful partnerships with commercial and academic institutions.

As we embark on this journey, we are committed to building a safe and sustainable future for GOSH and paediatric healthcare. By focusing on transparency, ethical compliance and continuous evaluation, we will maintain patient and staff confidence in our AI initiatives and continue our legacy of innovation to transform the lives of children and young people with rare and complex diseases.

Artificial Intelligence (AI) simulates human intelligence in machines programmed to think and learn like humans. It includes technologies like machine learning, natural language processing, computer vision and robotics. AI systems analyse vast amounts of data, recognise patterns and make decisions with minimal human intervention. Appendix 3 provides further details on the different types of artificial intelligence. For Great Ormond Street Hospital for Children (GOSH), AI has the potential to enhance patient care, streamline hospital operations and automating routine tasks, allowing healthcare professionals to focus on more complex aspects of patient care and support medical research by providing new insights.

The rapid advancement of AI presents an unprecedented opportunity for GOSH to revolutionise healthcare delivery. This strategy document outlines our vision to harness AI's potential to enhance clinical outcomes, drive efficiencies, and play a pivotal role in our future hospital strategy, including the building of a new state of the art children's cancer centre. By integrating AI into our operations, we aim to improve patient care, reduce clinician burden, and optimise hospital efficiency in a safe and sustainable manner.

At GOSH, we are committed to pioneering innovative solutions that transform healthcare and related research. Our AI strategy is designed to align with both the broader Trust strategy and national strategy and directives, ensuring that AI initiatives support clinical, operational and financial improvements.

To safeguard data security, GOSH has restricted access to freely available AI tools on the internet such as Google Gemini, ChatGPT and the free web-based version of Microsoft CoPilot. This is essential to prevent sensitive patient, staff or commercially sensitive information from being inadvertently shared online. Our next steps in AI will focus on providing staff with the right tools to enhance efficiency and patient outcomes.

Implementing AI-driven solutions will enable GOSH to increase the degree of personalised treatment we are able to provide, enhance clinical decision making and improve resource management. By building a scalable AI infrastructure and fostering both commercial partnerships and partnerships with academic institutions and other NHS organisations, we will develop a skilled AI team to drive this transformation. Importantly, we will also upskill and educate our wider workforce on safe and ethical AI use.

The integration of AI into GOSH's operations offers numerous benefits that will significantly enhance our healthcare delivery. AI can support diagnosis by providing AI-assisted diagnostics, which enhance clinical decision making and reduce diagnostic errors. By streamlining administrative tasks, AI drives efficiencies and productivity gains, allowing clinicians to focus more on patient care. AI can also optimise hospital operations through AI driven scheduling, resource planning, and cost efficiency, ensuring that our resources are used effectively.

Data and analytics will underpin evidence-based decisions to enhance care for our children and young people while driving organisational efficiencies. AI will be central to our transformation into a data-driven organisation.

GOSH will leverage its in-house expertise in digital technology, data and analytics, clinical system development and our DRIVE innovation unit. This expertise will be key in developing and implementing AI-driven solutions that enhance patient care and operational efficiency.

Our vision is to deliver real world AI solutions that improve patient outcomes, our corporate functions and the working lives of our staff in a safe and sustainable way.

We recognise the challenges of AI adoption including trust issues, data security and the need for human oversight. Our strategy focuses on transparency, ethical compliance and continuous evaluation to maintain patient and staff confidence.

Deploying AI at GOSH involves addressing concerns about accuracy, bias, data security and fears of AI takeover. Building trust through transparency, ethical compliance and continuous evaluation is crucial. Robust human oversight with comprehensive policies, governance frameworks and regular bias audits is essential. Investment is needed for technical infrastructure and expertise.

High quality data is essential for training reliable AI models. Healthcare data often suffers from inconsistencies, missing values and biases that can compromise AI-driven insights. Our strategy emphasises robust data governance frameworks and continuous data quality assessments. We will implement stringent data validation processes and advanced data cleaning techniques to ensure dataset integrity. Regular bias audits will identify and mitigate disparities in the data, ensuring our AI systems provide equitable and accurate outcomes. By fostering a culture of data excellence and integrating comprehensive data management practices, we aim to build a reliable AI infrastructure that enhances clinical decision making and operational efficiency.

Addressing these challenges will enable GOSH to enhance clinical outcomes, drive efficiencies and improve patient care. We are committed to building a safe and sustainable future for paediatric healthcare.

To deliver our AI vision for GOSH, we will focus on three strategic priority areas underpinned by three strategic enablers.

3 priorities:

• Enhance clinical decision making and care with AI
• Streamline operations and enhance efficiency
• Optimise non-clinical workflows

3 enablers:

• Build AI infrastructure and expertise
• Create AI governance and trust
• Drive collaboration, research and innovation

By focusing on these three strategic priority areas and three strategic enablers, GOSH can harness the power of AI to enhance clinical outcomes, streamline operations and drive innovation in paediatric healthcare. This comprehensive approach will ensure that AI is integrated safely and effectively, ultimately improving the quality of care for patients and supporting GOSH’s mission to pioneer innovative digital solutions in healthcare.

Implementing AI-assisted diagnostics will support clinicians in making more accurate and timely decisions. This includes:

• AI-powered imaging: Using advanced imaging techniques to detect anomalies and provide detailed insights that aid in diagnosis.
• Predictive analytics for early disease detection: Leveraging data to identify patterns and predict potential health issues before they become critical.
• Personalised treatment recommendations: Using AI to analyse patient data and suggest tailored treatment plans that improve outcomes.
• Enhanced patient monitoring: Providing real-time alerts to ensure timely interventions.

Integrating AI with our Epic Electronic Patient Record (EPR) will streamline workflows and reduce administrative burdens, allowing clinicians to focus more on patient care. We will use the AI solutions available within Epic, including the implementation of Epic’s Nebula cloud platform. This platform provides a scalable and secure environment for deploying machine learning models, enabling data-driven insights and improved patient care. Nebula empowers Epic sites to unlock the full potential of the system, including pre-built Epic models and custom algorithms developed to address specific clinical challenges, driving innovation, improving patient outcomes, and enhancing operational efficiency.

AI-driven solutions can greatly enhance resource planning, scheduling and cost efficiency. Machine learning and advanced analytics add significant value across the organisation, benefiting both clinical and non-clinical colleagues. In 2024, GOSH took part in a London-wide pilot of ambient voice technology AI (AVT) to improve clinician support during patient consultations.

• 240,000 outpatient appointments at GOSH in 2024
• 5,000 patient pilot of AVT across London
• 100 patient pilot at GOSH
• 10% reduction in consultation time with AVT

Data from the pilot indicates an average 10% reduction in consultation time per appointment. With GOSH clinicians handling 240,000 outpatient appointments in 2024, this efficiency gain translates into significant operational savings. Achieving a 10% efficiency improvement across all outpatient consultations could yield substantial benefits for GOSH.

Further planned initiatives include:

• AI-based staff shift planning: Optimising staff schedules to ensure adequate coverage and reduce burnout.
• Surgery scheduling: Using AI to manage operating room schedules, minimising downtime and maximising efficiency.
• Predictive models for ICU demand and bed occupancy: Forecasting patient needs to ensure resources are available when and where they are needed.
• Ambient AI technology: Utilising ambient AI technology during clinical consultations can significantly enhance efficiency and patient care. This technology allows clinicians to spend more time with patients rather than typing on a computer, improving the overall consultation experience and fostering better clinician-patient relationships.
• Provide summaries of care ahead of clinic visits and ward rounds.
• Provide suggested responses to messages from patients.
• Tailor order selection to condition type.
• Automate pathway steps such as adding to waiting list from clinic notes.
• Collate information into discharge summaries reducing manual creation.

Reducing the administrative burden on staff through automation will free up time for clinicians to focus on patient care. In 2025, GOSH will pilot CoPilot technology to streamline administrative workflows. CoPilot has created efficiencies in various sectors, including health and public sectors.

A 2024 CoPilot pilot at Buckingham Country Council found significant efficiency improvements and cost savings.

• 2024 Coilot pilot at Buckingham Council
• Pilot users between 10-20% more efficient
• Admin saved 60-90 mins per day

Specifically, the Council reported time savings of 10-20% for staff using CoPilot in business services. This reduction in time spent on routine tasks has allowed staff to focus on more meaningful work, thereby enhancing overall productivity and work quality. Within Customer Services it was found that customer services agents were able to save an average of 2 minutes per call which equates to a 33% efficiency saving. Administrators taking part in the pilot reported a 20% reduction in task completion time with an average of 60-90 minutes each per day. Generative AI has the potential to transform administrative workflows at GOSH by automating routine tasks such as data entry, document creation and billing and improving overall efficiency by handling tasks quickly and accurately, reducing errors and delays.

Further planned initiatives include:

• Automate pathways so that Referral-To-Treatment (RTT) pathways are closed when patients are discharged back to their GP.
• Provide predictions as to which patients are at a higher risk of not attending so reminder services can be more effectively targeted.
• Optical character recognition of scanned images and documents to enable quicker information collection for audits.

Developing a scalable AI infrastructure that integrates seamlessly with existing hospital systems is crucial. To achieve our vision, it is vital that we build reliable technology infrastructure platforms to support AI systems. This includes both cloud services and on- premise infrastructure with specialised hardware for AI development and deployment. Additionally, GOSH needs to invest in both the training and recruitment of a team of AI specialists, data scientists and clinicians with the expertise to develop and maintain AI solutions. GOSH will establish a cross-functional Centre of Excellence to oversee AI efforts. This team will be responsible for driving AI research, development and implementation across the hospital.

Implementing robust cybersecurity measures is vital to safeguard patient data and maintain trust in AI-driven healthcare services. We will establish stringent controls and continuous monitoring, using advanced AI-driven security tools to detect real-time anomalies and automate responses to evolving cyber threats beyond AI-specific risks. This proactive stance ensures our AI ecosystem remains secure and resilient.

To effectively use AI at GOSH, we must ensure our information is well understood and securely stored in the correct format and on the right platforms. Generative AI becomes more valuable as it gains knowledge and understanding of our organisation. Interoperability between our generative AI tools and the platforms that store and process our information is crucial.

To ensure the secure and responsible use of AI, GOSH will adopt the principles outlined in the AI Playbook (appendix 1) for the UK Government. These principles include using AI lawfully, ethically and responsibly, ensuring meaningful human control at the right stages, and managing the full AI life cycle. By adhering to these principles, GOSH can build a robust AI infrastructure that not only enhances clinical outcomes but also maintains public trust and complies with regulatory standards.

To build trust among patients, families and staff, it is essential to establish policies for the ethical and safe use of AI. This includes:

• Developing AI governance frameworks with polices for AI use within the hospital.
• Conducting regular bias audits to ensure AI systems are fair and unbiased.
• Engaging with key stakeholders (patients, families and staff) to address concerns and misconceptions about AI through transparent communication and involvement.
• Ensuring transparency by maintaining openness about AI processes, decision- making, and data usage to foster trust and confidence among all stakeholders.

AI must be transparent and explainable to patients, families and staff. Regular bias audits and fairness assessments must be conducted. Patient and staff engagement is essential to address concerns and build trust.

For safety and reliability, AI solutions must integrate with medical records for accountability. Establish AI revalidation and anomaly reporting systems; and ensure staff training on AI safety and oversight protocols.

Partnering with academic institutions, NHS Trusts, industry leaders and other stakeholders will drive AI research and innovation. Key actions include:

• Sharing best practice: Collaborating on AI projects that align with clinical, operational and financial needs, such as predictive analytics.
• Fostering innovation: Encouraging the development of new AI technologies and solutions to improve patient care and hospital operations, including AI in genomics and rare disease research for novel insights into paediatric genetic disorders and rare or complex diseases.
• Predictive analytics for disease: AI modelling for childhood disease outbreaks, clustering, prognosis and resource planning.
• Clinical trial optimisation: AI-driven patient recruitment and drug response prediction.

This strategy aligns with the UK government's three strategic shifts outlined in the “Road to recovery: the government’s 2025 mandate to NHS England”, published on 30 January 2025:

• From hospital to community
• From analogue to digital
• From sickness to prevention

By enhancing clinical decision-making and care with AI, the strategy supports the shift from hospital to community through more accurate and timely diagnoses, personalised treatment recommendations, and real-time patient monitoring. This can reduce hospital admissions and improve community-based care.

This strategy's focus on building a scalable AI infrastructure and integrating AI with existing systems like Epic aligns with the shift from analogue to digital, ensuring that GOSH leverages advanced technologies to streamline operations and enhance efficiency.

Additionally, the proactive use of AI for predictive analytics and early disease detection embodies the shift from reactive to proactive. This allows GOSH to anticipate and address health issues before they become critical, ultimately improving patient outcomes and operational efficiency.

The sustainable use of AI is integral to this strategy. We will prioritise energy efficient hardware and cloud services to minimise the environmental impact of our AI systems, selecting platforms and technologies that require less computational power to reduce energy consumption and environmental impact. Our AI development and deployment processes will adhere to sustainable practices, including working with suppliers and partners who prioritise renewable energy sources. We will focus on creating AI models that are both effective and resource efficient.

This strategy is designed to align with the regulatory compliance requirements and standards in the UK healthcare and research sectors. Ensuring adherence to these regulations is crucial for maintaining patient safety, data security, and ethical standards.

The GOSH AI strategy aligns with the EU AI Act (July 2024) by ensuring AI systems are transparent, accountable, and fair. AI systems will be robust, secure, and non-discriminatory, with human oversight to ensure clinicians have the final say. Data privacy and security will comply with GDPR, and an AI ethics committee will oversee ethical implications. Continuous monitoring will maintain high standards, ensuring AI technologies at GOSH are safe, ethical and enhance patient care.

Implementation of AI at GOSH will comply with the General Data Protection Regulation (GDPR) and the Data Protection Act 2018. This legislation mandates the protection of personal data and ensures that patient information is handled with the utmost care. The implementation will also comply with the common law duty of confidentiality, which exists in parallel with the legislation. The AI systems will incorporate robust data anonymisation and encryption techniques to safeguard patient data. Regular audits and assessments will be conducted to ensure ongoing compliance with the duty of confidentiality and data protection legislation.

AI solutions used in clinical settings will be classified as medical devices and will adhere to the Medical Device Regulations 2002 (MDR). This includes obtaining the necessary certifications and approvals from regulatory bodies such as the Medicines and Healthcare products Regulatory Agency (MHRA). GOSH will ensure that all AI-driven medical devices undergo rigorous testing and validation to meet safety and efficacy standards.

The ethical development and use of AI is central to GOSH's strategy. Our approach is built on transparency, accountability and fairness. To uphold these principles, we will establish an AI ethics committee. This committee will oversee all AI projects, ensuring they align with our core values and ethical standards. The ethical standards section is further expanded in Appendix 2.

To maintain high levels of patient care, we will establish clear protocols for AI use in clinical decision-making, ensure AI recommendations are subject to human oversight, and provide staff training on the safe and effective use of AI tools. Further information on the staff training plan is in appendix 7. Regular reviews and updates to the clinical governance framework will incorporate new developments in AI technology and regulatory requirements.

To ensure ongoing compliance with regulatory standards, GOSH will implement a continuous monitoring and evaluation process for its AI systems. This will involve regular performance assessments, bias audits, and impact evaluations to identify and address any issues that may arise. Feedback from patients, families, staff and regulatory bodies has been incorporated into the AI strategy to ensure it remains aligned with best practices and regulatory requirements.

GOSH will ensure that all partners and suppliers using AI adhere to the highest standards of safety, ethics, and regulatory compliance with an expectation that suppliers are transparent about their use of AI in processing GOSH data.

By aligning the AI strategy with these legal and regulatory compliance requirements and standards, GOSH will build a safe, ethical, and effective AI ecosystem that enhances patient care and operational efficiency. Continuous monitoring will include deploying AI management platforms such as Newton Tree to ensure all custom algorithms are assessed before deployment. A team of clinical scientists will robustly test the AI before, during, and after implementation to ensure it remains fit for purpose. All staff will receive training in their responsibilities in the appropriate use of AI, coupled with standards of practice, ensuring that in clinical settings, a record of what was originally produced by AI and how it was adapted is documented in the patient notes.

To fully leverage and deliver these technological advancements, a digitally confident and enabled workforce is essential. We recognise the varying levels of digital literacy among staff and will ensure support for all in using digital resources. A digital literacy assessment was undertaken at GOSH in 2024 to evaluate our current position across the Trust and identify knowledge and skill gaps, enabling us to create a comprehensive support and training package. We will collaborate with colleagues across the organisation to develop the digital workforce needed to fully exploit AI.

To deliver the AI strategy and vision at GOSH, we will use the “Artificial Intelligence Playbook for the UK Government” (February 2025) as our guide. This playbook outlines ten principles for the safe, responsible, and effective use of AI in government organisations. By following these principles, we will navigate AI deployment complexities, maximise AI benefits and mitigate risks. Further information on our approach to risk mitigation can be found in appendix 8.

Ensuring robust benefit realisation is essential. We will establish clear metrics and KPIs to measure the impact of AI initiatives, regularly review progress against these metrics, and make necessary adjustments to maximise benefits. This will involve continuous engagement with stakeholders to ensure that the AI solutions deliver tangible value and align with organisational goals.

In the first year, the focus is on laying the groundwork for the AI strategy. This involves defining who will lead the initiative and establishing the monitoring and reporting structures within the Trust. A gap analysis will be conducted to assess current capabilities, infrastructure, and use cases (appendix 6). Using the Healthcare Information and Management Systems Society (HIMSS) Analytics Maturity Assessment Model (AMAM) framework, a roadmap will be created to guide the development of an AI-driven hospital and assess AI readiness within the organisation. An AI governance structure will be established, including the formation of an analytics council.

The Department of Advanced Data Analytics and AI will be set up along with a governance framework. Pilot projects will be launched to test AI-driven solutions in diagnostics, scheduling, and administrative automation. AI literacy training will be developed for staff to ensure they are well-equipped to work with AI technologies. Continuous stakeholder engagement will be maintained to address any trust and perception concerns. Year 1 will also see the implementation of Nebula Proof of Concept within Epic.

During this phase, the focus will be on scaling and optimising AI use across various hospital operations. AI applications in diagnostics, operations and financial management will be expanded. An AI monitoring system will be developed to track performance and ensure safety. Partnerships will be formed with the NHS, industry, and academia to collaborate on healthcare AI research projects. We will advance towards HIMSS AMAM level 7, which signifies a high level of AI-driven hospital operations.

In the final phase, the goal is to achieve full integration of AI technologies and take a leadership role in the field. AI-driven precision medicine will be implemented at scale, allowing for more fully personalised and effective treatments. AI-powered robotic assistants will be deployed for paediatric surgeries, enhancing surgical precision and outcomes. The organisation will lead AI collaboration efforts across Integrated Care Systems (ICS), NHS Trusts, and academic networks, fostering innovation and knowledge sharing.

This strategy will significantly improve the experience and outcomes for our patients and their families. Ambient AI will streamline patient consultations, allowing clinicians to focus on patient interaction rather than data entry. AI-assisted diagnostics will support more accurate and timely decisions, leading to better-informed diagnoses and treatment plans. Highly personalised treatment recommendations will ensure each patient receives the most effective care tailored to their needs. Enhanced patient monitoring through AI will enable timely interventions, preventing complications and improving overall patient care.

This strategy is set to significantly enhance care delivery at GOSH. Ambient AI will reduce the administrative time clinicians spend during consultations, allowing them to focus more on patient care. AI-powered imaging and predictive analytics will enhance early disease detection, leading to better patient outcomes. Additionally, integrating AI with the Epic EPR will streamline workflows and minimise administrative burdens, making the entire process more efficient.

This strategy is set to significantly enhance support services. By leveraging generative AI in administration, we anticipate notable efficiency improvements. AI-driven solutions will enhance resource planning, scheduling and cost efficiency.

This strategy will enhance research by facilitating partnerships with academic institutions, NHS Trusts, and industry leaders, providing researchers with opportunities to collaborate on cutting-edge AI projects. Researchers will have access to AI-driven solutions that support diagnosis, streamline administrative tasks, and optimise hospital operations, allowing them to focus more on innovative research. AI will enable better data management through predictive analytics and personalised treatment plans, enhancing the quality and efficiency of research.

By fostering partnerships and creating a scalable AI infrastructure, the strategy will attract funding and resources, supporting researchers in their endeavours. These benefits will position GOSH at the forefront of healthcare innovation, setting new standards for paediatric care and ensuring a brighter future for patients and researchers.

AI will optimise hospital operations, improve resource management, and support medical research, ensuring a safe and sustainable future for paediatric healthcare.