Are you ready to scale AI?

The question facing business leaders today is no longer whether they can use AI, but how they can move from fragmented experiments to a scaled and strategic implementation with a measurable return, according to a panel of AI experts at the Global Risk Summit, sponsored by Airmic, last month.

While businesses are building many use cases, it is time to move beyond proof of concepts, argued Swiss Re’s group chief digital and technology officer, Pravina Ladva: “The next stage is thinking about how this technology can drive real business value and help solve industry problems that you’ve faced for many years.

“And that’s why it’s a CEO topic, it’s a board topic, and actually first and foremost, it’s a business topic.”

Scaling AI and driving value requires purpose, focus and intention, she said. Swiss Re has been through a process of “fundamentally reimagining” some of its core insurance processes in the context of AI – and with tangible results.

For example, in data-intensive lines of insurance like engineering and construction, they have used AI systems to reduce the time to offer from 45 days to just one. This required cross-functional collaboration: “In a room of about 15 people, there were probably only two technologists. The rest of them were business folk,” said Ladva.

Trust and governance: a boardroom conversation

Investment bank and financial services firm, Citi, has scaled its use of AI rapidly with over 190,000 of its global workforce using AI, representing an 80% adoption rate. Citi’s chief technical officer, David Griffiths, said building trust in their systems and processes, both internally and externally, has been essential.

Taking a methodical and well-communicated approach was important, he said, as well as embedding guardrails at the outset to minimise risks. “We have been very intentional about the technology design, how to implement the technology controls, how we explain those controls,” he said.

Having trust in your data is also vital for successfully scaling AI, the panel agreed. Vishal Marria, founder and CEO of Quantexa, a data, analytics and AI software business, believes the saying “garbage in and garbage out” rings true.

“The explainability and transparency of the data…is now a board level conversation,” he said. Furthermore, there needs to be “clear alignment” between the board and the operational teams: “If you do not have board alignment, these will end up as proof of concepts.”

Empowering not replacing the workforce

While few dispute the potential benefits of scaling AI across businesses, many remain concerned about the impact on workforces and finding the appropriate balance between efficiency and human judgement.

Marria believes all companies have an obligation to implement AI in a way that augments the use of human experience. “The human aspect is foundational. It’s not a sidebar responsibility; it’s at the core,” he said.

His company of approximately 1000 employees has achieved 80% AI adoption. While some processes operate with minimal human input, for the most part, “humans in the loop are key”.

Similarly at Citi, AI is framed as a workforce enabler. “One of our stated goals is to make our workforce feel like the most AI-empowered workforce in the world,” noted Griffiths.

While roles that are repeatable and codable may change, the human race does not sit still, he said, and concerns around AI are rooted in the fear of the unknown. “The fear is we don’t know what the future will look like – but there will be one. And the exciting thing is that we are currently defining what that will look like,” he said.

Nevertheless, as organisations move from isolated use cases to scaled AI, they have a responsibility to reassure their workforce, according to Swiss Re’s Ladva. If the technology is 20% of the AI challenge, then change management and people engagement accounts for about 60%.

This means proactively bringing people along on the journey: “You will see spikes [in AI engagement] and then it will plateau…so keeping that momentum and keeping that training and education is key,” she said.

Will you be on the right side of history?

Predicative artificial intelligence has been in existence for many years, but the advent of scalable generative AI in the past three years has “catapulted” its impact, according to the panel.

As the panel’s moderator, Max Richter, EMEA CEO at mea, said, we are entering the next phase of AI, where it “has moved from the innovation lab to the boardroom.” It is transforming productivity, resilience, regulation, workforce strategy – and above all, competitive advantage.

In the words of Marria: “We mustn’t as a society here in the UK push this away; we will lose competitive advantage if we do. So, we have to bring this in and to make it work for our people.”

Today, the biggest business risk of all, he warned, is being left behind: “This is a once in a generation technology. You’ve got to be on the right side of history.”

The AI debate took place the Global Risk Summit, sponsored by Airmic, and part of the inaugural London Risk Week, 11-15 May.