Beyond The Buzz: Liberty IT’s Journey To Responsible Generative AI

How Curiosity, Collaboration And Responsible Innovation Shape The Future Of AI

Generative AI has transformed the way businesses work, learn, and create - and at Liberty IT, we’ve seen that transformation first-hand. But behind the headlines and hype, what does it take for an organisation to truly harness its power responsibly, while keeping people at the heart of every innovation? 

At Liberty IT, curiosity and experimentation have always been part of our DNA. It’s how we stay ahead of emerging technology trends and continue to evolve as one of the leading digital innovation hubs supporting Liberty Mutual’s global transformation. 

We caught up with Sinead Shackley, Senior Director of Engineering at Liberty IT, to talk about how her teams are pioneering the use of Generative AI (GenAI), the lessons learned along the way, and why the company believes that AI may do the grunt work, but people will always do the great work. 

Leading With Curiosity And Innovation 

Sinead, for us at Liberty IT, the GenAI journey began long before ChatGPT. Could you tell us a bit more about this? 

At Liberty IT, we regularly scan for emerging technologies. We’d bring people together from across the company to explore new technologies and consider their relevance to our industry. 

Towards the end of 2022, we identified GenAI as an interesting technology and even created a white paper on it. To be honest, we didn’t move on it straight away, but when ChatGPT went mainstream in early 2023, we revisited our white paper - and that’s what started everything for us.

Would you say that the early curiosity paid off? Having already explored the technology’s potential and risks, did that mean we were able to move more quickly? 

It did. We were quickly able to stand up several foundational capabilities - including our own in-house version of ChatGPT. From there, we built the rest of the ecosystem: a GenAI architecture, reusable capabilities and identified use cases that aligned with our business needs.

Building An AI Ecosystem That’s Responsible By Design 

A lot of organisations are still figuring out how to balance innovation with responsibility. How did Liberty IT approach this? 

Innovation and responsibility go hand in hand. From the start, we knew we wanted a structure that ensured AI was used ethically and transparently. That’s why we developed a multi-step governance process for every Generative AI use case. 

We advocate for a ‘Responsible AI by Design’ process - embedding governance directly into product development. It means understanding potential risks and impacts early on, so we can innovate confidently and with integrity. 

Every AI initiative at Liberty IT is designed to drive innovation, but also to do so with accountability, transparency and trust at its core.

AI In Action 

Once the foundations were in place, how did we start to see AI make an impact? 

Honestly, the impact has been huge. Our in-house ChatGPT instance alone has saved hundreds of thousands of hours, giving people time back to focus on higher-value work. We also built an AI platform that supports a portfolio of foundational and frontier models - those are being used right across our business.

Our engineers use GitHub Copilot to accelerate coding, and we’ve applied GenAI to complex database conversions that once took months. In one project, we tested different models until we achieved 100% accuracy — saving significant external spend. But beyond efficiency, AI has helped us unlock so much hidden knowledge.

Hidden knowledge? Can you tell us more? 

We realised how much tacit knowledge was sitting across platforms like SharePoint or Slack - knowledge that wasn’t easily searchable. With AI-powered knowledge assistants, our engineers and claims teams can now surface that information in seconds. It’s been a real game-changer for collaboration and learning.

Overcoming Challenges – It Takes A Village 

Every big change comes with its challenges. What were some of the biggest hurdles the team faced? 

Like any organisation, we had plenty of hurdles — especially at the start. Limited access to models, throughput issues, even applying for extra tokens to keep our experiments going! 

But honestly, the biggest challenges weren’t technical — they were about people, process, and data. It really took a village. We worked closely with colleagues from privacy, compliance, data science, security, infrastructure, and procurement.

Everyone played a role in making this work. That cross-functional teamwork is what made it possible to move quickly and safely.

There’s always debate about whether AI replaces or supports human roles. What’s your perspective on this? 

For us, AI is here to enhance human capability, not replace it. The trend we’re seeing — and the one we embrace — is a hybrid model. AI does the grunt work, humans do the great work. In practice, that means engineers use AI to accelerate coding and documentation, but it’s people who lead on design, problem-solving and decision-making. 

In customer operations, for example, AI can summarise information or suggest insights, but it’s humans who bring empathy, nuance and judgment. AI reshapes roles more than it replaces them - and that’s where adaptability and continuous learning become so valuable.

Shaping People’s Skills 

How are we helping our people grow in this space? 

GenAI is transforming how we work and serve our customers, so we’re focused on giving everyone the skills and confidence to experiment responsibly. That’s where our GenAI Learning Mission comes in — it’s designed for everyone, from engineers to people leaders. 

The initiative includes multiple learning paths: from AI literacy and ways of working, to building AI into applications and evolving technical skills. We also run monthly office hours with our experts, along with promptathons, unconferences and knowledge shares. 

It keeps everyone connected and learning from each other.

Has Generative AI changed the skills Liberty IT looks for? 

We’ve put more focus on critical thinking - because even as AI advances, the need for human oversight, creativity and judgment is stronger than ever.

Lessons For The Future 

If you had to sum up Liberty IT’s journey with AI so far, what would you say? 

Breaking the barriers to AI isn’t about the tech - it’s about the people. It’s about curiosity, collaboration, and building responsibly. Our goal is to make sure every AI initiative not only drives innovation but does so with accountability, transparency and trust at its core.

The belief that people remain at the heart of progress continues to define Liberty IT’s approach to AI and every new technology we explore. Interested in joining a team where innovation is driven by curiosity, collaboration and purpose? Explore careers at Liberty IT and help shape what’s next.