Davos is an opportunity for world leaders to explore insurance as risk management: MMC's Klint

When global leaders gather in Davos this week for the annual meeting of the World Economic Forum (WEF), they should go below the surface of what they believe the re/insurance sector can offer in their search for solutions to emerging risks, including those posed by artificial intelligence (AI), according to Carolina Klint, chief commercial officer for Europe at Marsh McLennan.

In an interview with Insurance Day, Klint says business and government decision-makers should leverage the expertise and knowledge of the sector “a bit better” than they do at present.

“If they scratched the surface, they would find a lot of competence, data aggregation models and knowledge that could be beneficial for them on a much broader scale,” Klint says.

Along with Zurich Insurance Group, Marsh McLennan collaborated with the WEF on its Global Risks Report for 2024.

Notably, “adverse outcomes of AI technologies” is a new entrant to the top 10 list of global risks ranked by severity. This risk is fairly low down in the two-year outlook – 29th out of the 34 risks – but it jumps to number six in the 10-year outlook.

AI concerns

Alongside the possibility of an entity achieving “artificial general intelligence” – learning to accomplish any human or animal task – key concerns about AI cited by respondents to the report included: misinformation and disinformation; job loss and displacement; criminal use and cyber attacks; bias and discrimination; use in critical decision-making by both organisations and states; and AI’s integration into weaponry and warfare.

The elevation of re/insurance to discussions among public and private sector leaders is an “organic development”, Klint says. Their approach to risk must not be limited to risk transfer, but also include risk identification, risk mitigation and risk management, she continues. She emphasises these are all part of the skill set of re/insurers.

There is already evidence, she adds, the industry is being seen as “so incredibly relevant” to addressing global risks. “What I’ve definitely seen is risk as a topic and risk connected with strategy have become an agenda point for boards, C-suite members, for much more senior leaders and stakeholders than we’ve seen in the past. And I think there’s a recognition, in this complex and interconnected business environment, you must connect risk with strategy to build a business model that is resilient and will take you into the future,” she says.

But this trend is a two-way street. In terms of what Marsh has gained from its collaboration with the WEF on its Global Risks Report, Klint says “perspective on the risks facing humanity, society and business”.

She adds: “It is such an incredible opportunity to get a holistic view of the risk landscape, from the coming together of all these experts from different areas and representing different functions. Such diversity is really the only way you can get a holistic view.”

Klint spoke at the launch of the report alongside John Scott, head of sustainability risk at Zurich Insurance Group, and WEF managing director, Saadia Zahidi.

“AI-generated misinformation, cyber attacks and cyber insecurity have emerged as top risks across all time horizons,” Klint told reporters at the launch event. “And it’s about time, if I may say so, because these risks have not been prominently featured in past years, which I think is surprising given our dependency on techno­logy, but also given the acceleration of digitisation following the pandemic.”

The “sudden accessibility” of user-friendly interfaces like ChatGPT, larger-scale AI models and synthetic content to manipulate public opinion means “it is no wonder” misinformation and disinformation is ranked by respondents to the report as the top risk in the short-term outlook, she says.

“The potential impact on elections worldwide over the next two years is significant. And that could lead to elected governments’ legitimacy being put in question and this, in turn, could of course threaten democratic processes leading to further social polarisation, riots, strikes or even intrastate violence,” she adds.

Opportunities and implications

Klint stresses the use of AI “comes with so many fantastic opportunities”, but businesses have to also consider the societal, economic and security implications of using it. And the success of this, she adds, will depend on “regular testing, updating verification methods and lots of upskilling and training”.

“It’s the only way we can prevent strategic and tactical miscalculations by AI that could lead to catastrophic errors, reputational harm or even significant liabilities,” she says.

AI provides bad actors with new tools and pathways to exploit, she continues. “In fact, you don’t even have to be smart nowadays to be a successful cyber criminal,” Klint says. “AI can create advanced malware, it can impersonate others for scams, it can poison AI training data, which is a very serious concern.”

On the other hand, AI can also be used for better responses to cyber attacks, she stresses, in that it can “process vast amounts of data, detecting patterns humans may overlook”.

The main feature of AI, used for good or ill, is speed. “Cyber criminals have made developments that are really faster than our ability to imagine and come up with the solutions for how to fight back. And with AI that risk is multiplying,” Klint says.

She illustrates this risk with the robbery of Bangladesh Bank in 2016, when cyber criminals stole $81m.

“They’d sent out 36 phishing emails to employees at the beginning of 2015. Three were clicked on and they got into the system. It took them a year to map out the security and get past the network, to figure out how it was wired. Had that been powered by AI, it would have taken two days,” she says.

“But on the other hand, the same is true for the response. It took months and months to try and backtrack to figure out what had happened and how to stop it [happening again]. With AI, it could have taken a couple of days. So again, we can use it for good, but it can also be used to really disrupt and put stress on our systems.”

AI can also automate many security processes such as patch management to fix bugs and close security holes, she continues.

“But we must recognise the fact everything we use, such as water, electricity, the financial system, the communication systems, is dependent on the integrity of incredibly complex networked computer systems, so cyber security is not simply about protecting a computer or protecting a file; it’s more about making sure supply chains work and society as a whole stays up and running,” she says.

In her conversations with companies to produce the WEF report, two issues consistently come up as key concerns, Klint says. One is the adverse impacts of AI and the other is supply chain challenges.

Supply chain challenges

On the latter, she points to recent positive news a brand-new material found using AI and super­computing could reduce reliance on lithium for batteries by up to 70%.

The findings were made by Microsoft and the Pacific Northwest National Laboratory, which is part of the US Department of Energy. They narrowed down 32 million potential inorganic materials to 18 promising candidates in less than a week in a screening process that could have taken more than two decades to carry out using traditional laboratory methods.

“Those kinds of advances – to reduce the need for precious metals or other things that can be difficult to come by – will also change the way we think about raw materials and the things we need,” Klint says.

“At this pivotal moment in the technological revolution, businesses have to take a very proactive approach to assess vulnerabilities, stay informed about regulatory developments and implement robust cyber security and risk mitigation measures,” she adds.

Misinformation and disinformation form a significant risk, she continued, and accessibility to large-scale AI models has made it “so much easier” to reach a large population of voters with content that “looks and feels real, even though it is not”. “It is the reason we see misinformation and disinformation being ranked as the top global risk in the short term, because there’s a level of awareness, but I’m not really sure we know how to get our arms around it and how to combat it,” Klint says.

“At this pivotal moment in the technological revolution, businesses have to take a very proactive approach to assess vulnerabilities, stay informed about regulatory developments and implement robust cyber security and risk mitigation measures”
_{Carolina Klint
Marsh McLennan}

On international and domestic actors looking to leverage AI to produce content that will sway public opinion and interfere with political election results, she says this threat is “probably coming from both nation states, but also internally and domestically”.

Re/insurers should really think about how to enhance their business models with AI, but still have humans verify the accuracy and contextual relevance of AI outputs.

“There are so many opportunities from AI, but it’s important to recognise we cannot lean on it completely,” Klint says. “It’s about upskilling the workforce to equip and empower them to be very efficient with AI as their copilot.”

She continues: “AI will create more winners and losers. It will drive a wedge between the companies that focus on the education and the recruitment of the right talent for it – and focus on the risk management of AI as well – and the companies that don’t.

It is crucial, she says, to pay attention to the risk of cyber attacks aimed at infecting AI training data and introducing bias into data models. “When that happens, it’s incredibly difficult to discover, and then, once you’ve discovered it, it’s incredibly difficult to clean it up.”

AI’s jump into the top 10 risks listed in the WEF report reflects uncertainty around approaches to regulation, Klint stresses.

“The EU is now issuing its AI Act, but once it is enforced, it might well be superseded by the extra­ordinary pace of AI’s evolution. It’s that pace which is making it so difficult as a risk.”

Insurance Day will be publishing a special report next month on market modernisation, including the topic of artificial intelligence.