Leading into Tomorrow

In the final episode of AI Pioneers, EY Global Government and Infrastructure AI Co-leaders Sheena Thompson and Thomas Erwin share their highlights from across the miniseries.

They identify four recurring themes:

1. Successful AI programs begin by understanding the problem, not the tool.
2. Adoption, productivity, and ROI depend on training your people and bringing them with you.
3. Radical transparency and giving citizens control earn public trust.
4. Standardized governance, built for scale, doesn’t block innovation; it operationalizes it.

Sheena and Thomas also provide a taste of the insights their guests shared under each theme. For more detail, dive into the other episodes.

This is the eighth episode of AI Pioneers, a miniseries featuring speakers from pioneering government organizations identified in EY data and AI reports. Listen in for practical lessons from peers who have successfully moved from ambition to impact – scaling AI to transform their organizations and how they serve citizens.  22 mins

In this episode, EY Global Government and Infrastructure Leaders, Thomas Erwin and Josie Cluer, discuss a key theme in EY data and AI reports: workforce engagement and capability. Josie shares some findings from global EY research:

• Less than half of government employees are seeing productivity benefits from AI, compared with over three quarters of employees in financial services.
• Skills are twice as effective as tools or mindset in helping employees realize productivity benefits from AI.
• The more training people do, the bigger the gains.

The discussion also covers the importance of change management. Josie cautions that while government should be ambitious about AI’s potential, moving fast without addressing employee concerns risks leaving people behind.

This is the seventh episode of AI Pioneers, a miniseries featuring insights from pioneering government organizations identified in EY data and AI reports. Listen in for practical lessons from peers who have successfully moved from ambition to impact – scaling AI to transform their organizations and how they serve citizens. 18 mins

From 10 use cases to almost 100 at the proof-of-concept stage or in production: the Île-de-France Region has scaled GenAI fast. Now, it’s accelerating its use of agentic AI to improve everything from employee productivity to the citizen experience.

In this episode, Chief Information and Technology Officer Achille Lerpenière tells Sheena Thompson, EY Government and Infrastructure AI Co-leader, how a hybrid operating model enabled this success. Encouraging employees to try out the technologies, rather than mandating their use, also built skills and trust. With Île-de-France now considering tools for governing and securing agentic AI, Achille highlights the need to prepare for what’s coming next – including new risks from quantum computing.

This is the sixth episode of AI Pioneers, a miniseries featuring speakers from pioneering government organizations identified in EY data and AI reports. Listen in for practical lessons from peers who have successfully moved from ambition to impact – scaling AI to transform their organizations and how they serve citizens.  14 mins

When the City of Amarillo, Texas, needed to service a rapidly growing population that couldn’t afford higher taxes, it was the catalyst for exploring how AI and automation could help.

As Chief Information Officer Rich Gagnon tells Sheena Thompson, EY Government and Infrastructure AI Co-leader, a key solution was creating a GenAI concierge, Emma, to handle website requests. But with a large multilingual population, and low initial levels of digital literacy, trust had to come first. That’s why establishing a clear digital dignity, privacy and transparency agenda, and co-designing and testing the tool with residents, were foundational steps.

Today, Emma reflects the language needs and usage preferences of the community while handling around a third of all service requests and deferring around US$1.8m a year in call-center hiring.

This is the fifth episode of AI Pioneers, a miniseries featuring speakers from pioneering government organizations identified in EY data and AI reports. Listen in for practical lessons from peers who have successfully moved from ambition to impact – scaling AI to transform their organizations and how they serve citizens.  17 mins

Governments should look for the best solution to a specific problem, not view AI as the answer to everything.  That’s one of the main lessons shared by Dr Laura Gilbert CBE, previously Director of the Incubator for AI at No 10 Downing Street. Laura tells Sheena Thompson, Government and Infrastructure AI Co-leader, how the Incubator built trust through “competency and intent”. That meant sharing code, evaluations and failures, along with use cases to show it was acting in citizens’ best interests.

The conversation also covers the ethical implications of LLMs that are trained on data informed by local cultural norms. And Laura shares her aim of using AI to make the experience of government more human.

This is the fourth episode of AI Pioneers, a miniseries featuring speakers from pioneering government organizations identified in EY’s data and AI reports. Listen in for practical lessons from peers who have successfully moved from ambition to impact – scaling AI to transform their organizations and how they serve citizens. 20 mins

At the South African Revenue Service (SARS), AI-powered fraud and risk detection have stopped 100 billion rand in impermissible refunds and recovered 20 billion rand in revenue by dismantling illicit schemes. SARS also uses third-party data to auto-assess almost six million personal income tax returns, with a 99% acceptance rate.

Carl Scholtz, Deputy Commissioner, Strategy Enablement and Modernisation at SARS, tells Thomas Erwin, EY Global Government and Infrastructure Co-leader, how his organization designs digital services so that “tax just happens”. It also combines experimentation with world-leading governance and an ethical-by-design approach – for example, testing AI bot responses for relevance rather than deterministic consistency. The North Star: making sure taxpayers and employees have a great experience.

This is the third episode of AI Pioneers, a miniseries featuring speakers from pioneering government organizations identified in EY’s data and AI reports. Listen in for practical lessons from peers who have successfully moved from ambition to impact – scaling AI to transform their organizations and how they serve citizens. 26 mins

For Shen Hwee Chua, Chief Data Officer at the Energy Market Authority in Singapore, the biggest hurdle to implementing a GenAI productivity tool has been human resistance.

As she tells Thomas Erwin, EY Global Government and Infrastructure AI Co-leader, hosting the organization’s GenAI platform on a government-approved cloud was key, as was democratizing access and skills. Codesigning the tool with important stakeholders and continually gathering their feedback gained buy-in. And automating the process of making documents machine-readable saved many hours of manual effort.

In future, Shen Hwee believes AI technologies will not only make governments more efficient but also allow them to collaborate more dynamically – across departments, agencies and borders.

This is the second episode of AI Pioneers, a miniseries featuring speakers from pioneering government organizations identified in EY’s data and AI reports. Listen in for practical lessons from peers who have successfully moved from ambition to impact – scaling AI to transform their organizations and how they serve citizens. 16 mins

Not only do over half of all Estonians use AI daily, but on average, government employees are more likely than private sector employees to do so (37% v 25%). 

In this episode, Ott Velsberg, Estonia’s Government Chief Data Officer, explains how mandatory training, role-specific competencies and positioning AI as a productivity tool have normalized its use across the public sector. He tells Thomas Erwin, EY Government and Infrastructure AI Co-leader, how the government has put real-world experimentation before drafting policy and reformed AI funding to support testing hypotheses, not delivering outcomes. And he sets out how being transparent about how data is used and algorithms work can help build citizen trust.

This is the first in a miniseries featuring speakers from pioneering government organizations.  22 mins