Liverpool - People's AI Stewardship Summit

The Importance of Public Voice in AI with Eleanor O'Keeffe 

Eleanor O’Keeffe emphasised the urgency of public deliberation in AI policy, especially given the rapid pace of technological and regulatory change.

Rather than debating why public input matters, she focused on how practitioners can make it matter. The Ada Lovelace Institute integrates public attitudes, lived experience, and deliberative reasoning with legal and policy analysis to build stronger, evidence-based recommendations. A positive case study is the Citizens’ Biometric Council, which explored public concerns about technologies like face recognition concurrent with a review of governance and policy gaps. Together, these efforts directly influenced the EU’s AI Act.

Download a visual summary

Participants discussed how government, industry, and civil society could ensure AI serves the public good, ensuring benefits while minimising risks.

AI in Healthcare: Strengthening Oversight and Trust

Participants advocated for an independent regulatory body with real authority, alongside an organisation to oversee intellectual property rights and tighter legislation to protect patient data.

Regulation must protect both consumers and patients.

Participants called for a cohesive approach to policy and funding across government agencies to prevent fragmented decision-making.

Transparency and public trust were central. One idea was to widely publicise success stories of AI in the NHS to build confidence in its benefits.

There was a consensus that AI should be a tool to assist healthcare professionals, not replace them.

AI in Transport: Building a Fair and Safe System

A key concern in transport was ensuring equitable access to AI-driven advancements. Participants stressed that public benefit should take precedence over profit, echoing the need for regulations and rigorous testing before AI systems were widely deployed.

Public engagement was a central theme. Participants emphasised the need for open discussions, town hall meetings, and participatory decision-making.

Finally, maintaining individual choice was a priority, with calls for opt-out options for AI-driven transport services.

Five Enterprise Hub members held focus group discussions providing an opportunity for them to hear what the public felt about their specific AI applications. Later they shared some of the insights they received with the room.

Healthcare (ADHD Diagnosis)

While appreciating AI’s potential to improve ADHD diagnosis rates, participants were concerned about bias. Unease about data privacy, equitable access, and a lack of empathy were also voiced.

Emergency Services (Drones)

Significant enthusiasm for AI-powered drones in emergency response was tempered by ethical reservations about AI making life-or-death decisions. Participants felt there is a need for human oversight and careful integration with existing emergency response systems.

Sports Coaching (Performance Analysis)

Although AI tools were regarded as beneficial for athlete performance, participants stressed the irreplaceable value of human aspects of coaching, like an ability to notice emotions.

Education (AI and VR in Learning)

Participants discussed the need for AI literacy programmes to equip future generations with the skills to be involved in building AI systems and use AI responsibly.

Accountability (AI Across Sectors)

Participants expressed concern about bias in AI-driven decision-making and called for clear lines of accountability when AI systems produce unfair outcomes. 

How can we ensure that the benefits from AI developed in the Northwest and using local resident data have a visible and tangible impact on local communities?

A primary concern was the unclear ownership of local data. Participants proposed establishing cooperative data trusts, allowing residents to collectively manage and benefit from their data. A suggestion for local levies on AI developers was also put forward.

Existing local authorities, town hall meetings and parish councils could play a role in facilitating democratic engagement, letting residents speak up about how their data is used by, say, Mersey Transport or the local NHS.

Focusing specifically on AI in health and care, what sort of help or information would be needed to define other patients’ needs and ensure that AI-driven health and care services meet the community’s specific requirements?

Participants stressed the need for faster diagnoses, particularly noting that lengthy waits for scan results can exacerbate health problems, especially in elderly patients. Many expressed a desire for AI to provide more personalised care, which some felt is currently lacking due to limited data sets in health research.

Additionally, they discussed how AI could enhance social care, pointing out its potential to help identify and prevent critical issues like child criminal exploitation.

Considering AI’s broader applications, what other opportunities or problems in the Liverpool City Region could AI help to address, what are the priorities, and why?

Participants expressed interest in using AI to solve challenges like the energy crisis, advancing fusion technologies. They discussed the potential for AI to improve public services, optimising travel routes or identifying whether building sites are at risk of flooding.

The group considered whether AI could or should be directly involved in moderating balanced views on global issues and politics.
They also expressed concerns over selecting trustworthy AI systems; could brand recognition, peer reviews, and tools like Trustpilot play a role in establishing confidence?