Why AI is not the new blockchain
Learning from past technology projects
In 2018, Lloyd’s predicted that blockchain was about to transform the market. The three-hundred-year-old institution threw itself into the pursuit of blockchain-enabled transformation. As did a handful of big-name carriers (more on that below). Several million pounds and some hefty consultancy fees later, however, initial high hopes fell back to earth with a bump. Lloyd’s abandoned its plans for end-to-end policy management via blockchain and quietly wrote the investment off. The market moved on.
The blockchain bubble was just the latest in a long line of ambitious, but ultimately ill-fated digital initiatives in this market, joining the likes of Electronic Placing Support (EPS) and its successor EPS 2, Kinnect, and the Lloyd’s Exchange. All of which promised much, but delivered little.
Given this history, more skeptical commentators could be forgiven for suspecting that current flavour of the month, AI, could go the same way. But AI is different. For one thing, the use of AI within the London Market isn’t being imposed from the top down. Instead, we’re seeing organic localised adoption by those who recognise its potential to help with immediate challenges.
The market’s attempts to leverage blockchain depended on a radical structural shift imposed from above. AI adoption, by contrast, works within existing frameworks, enhancing current processes not replacing them
Why previous initiatives failed
The London Market has never been short of big ideas. But most have never found their way into practice. One reason so many technology-driven initiatives have failed is that some of those involved may not have fully appreciated exactly how the market works in practice.
Lloyd’s and its peers operate in a world of intricate negotiations, bespoke contracts and relationship-driven decision-making. Past projects have often assumed that efficiency could be imposed through rigid structures, ignoring the complexity and flexibility the market requires to function as it does.
EPS and Kinnect failed precisely because they underestimated the complexities of risk placement. Lloyd’s Exchange suffered from poor integration and a lack of clear incentives for adoption.
Then there’s the issue of buy-in. Unlike a more conventional corporation with centralised decision-making, the London Market consists of many alternately competing and cooperating businesses, each with its own technology stack and its own strategic priorities.
For a new system to succeed in London, it needs to allow for all these competing interests.
Why AI is different
AI will thrive where past initiatives haven’t because it doesn’t require a market-wide reset. It doesn’t replace human expertise. It amplifies it.
In claims handling, for instance, AI-powered triage systems can sift through incoming claims, flagging the ones that require urgent attention. This allows claims handlers to focus on complex cases rather than ending up buried in unproductive admin tasks.
Fraud detection is another area where AI is already proving itself invaluable. Machine learning models can identify suspicious patterns far more effectively than traditional methods, helping insurers reduce fraudulent payouts without adding layers of bureaucracy.
And then there’s natural language processing (NLP), which is transforming how insurers handle documents. AI can extract relevant information from loss notifications, policy wordings and medical reports, reducing reliance on manual data entry and improving accuracy.
AI adoption isn’t about forcing change. It’s about making existing systems work better. AI can integrate seamlessly with the market’s existing workflows, rather than needing to replace them.
Lessons from other industries
Insurance isn’t the only sector grappling with technological change. AI’s impact is being felt across industries, often in ways that provide useful parallels for the London Market.
Take radiology. AI isn’t replacing radiologists, but helping them work more efficiently. AI-powered imaging tools highlight anomalies in scans, allowing doctors to focus on the most complex cases. This drives faster diagnoses, greater accuracy and, ultimately, better patient outcomes.
The financial services sector offers another useful comparison. AI-driven risk assessment tools are helping banks detect money laundering and fraudulent transactions more quickly and accurately than ever before. Again, these tools aren’t replacing human expertise, but enhancing it – enabling compliance teams to focus on the highest-risk cases.
Legal firms, too, are finding ways to use AI to improve efficiency. For example, contract analysis tools scan legal documents for inconsistencies or missing clauses, freeing up lawyers to focus on more strategic tasks.
In each of these cases, AI operates as a force multiplier. It doesn’t eliminate the need for skilled professionals. It simply makes them more effective.
Top-down vs. market-led innovation
One of the biggest differences between AI and previous insurance technology failures is that AI is being driven by market needs, not mandates from on high. The insurance industry has seen plenty of top-down initiatives that failed because they didn’t align with how brokers, underwriters and claims handlers actually work. AI, by contrast, is being adopted because it delivers tangible benefits.
This bottom-up approach is critical. Insurers aren’t being told to use AI. They’re choosing to do so because it makes their lives easier. Underwriters are using AI-driven analytics because they provide better insights. Claims teams are embracing AI-powered automation because it reduces paperwork and speeds up settlements.
This kind of organic adoption ensures that AI doesn’t become another expensive white elephant. Unlike Kinnect or Lloyd’s Exchange, AI isn’t trying to change how the market operates. It’s simply making existing processes better.
The reality of blockchain in insurance
While AI is gaining traction, blockchain has largely failed to deliver on its early promise. For several years, blockchain was hyped as the future of insurance, particularly for automating claims payments and reducing fraud. Yet, despite significant investment, real-world adoption has been distinctively limited.
The Blockchain Insurance Industry Initiative (B3i) was launched with much fanfare in 2016, backed by major players like Allianz, Swiss Re, and Zurich. The goal was to create a distributed ledger system for catastrophe excess-of-loss coverage. But by 2022, B3i had filed for insolvency. Even blockchain’s biggest champions struggled to make it work.
The reasons are simple. Blockchain solutions require industry-wide standardisation, which is difficult to achieve in a sector as diverse and multifarious as (re)insurance. They also depend on clean, reliable data — something that remains a challenge to this day. And while blockchain’s transparency is appealing, it doesn’t yet offer enough practical advantages to justify the cost and complexity of implementation.
Even the firms that invested most heavily in distributed ledger technology are now focusing their efforts elsewhere. Swiss Re, once a blockchain pioneer, has shifted its attention to AI.
AI’s role in the future of insurance
Of course, blockchain is far from dead and buried. It still has a role to play in niche areas like parametric insurance and certain types of fraud prevention. But, for most (re)insurers, AI offers far greater potential. AI-driven claims automation, underwriting analytics and fraud detection are already delivering tangible benefits. And, because AI works with existing workflows rather than against them, the pace of adoption is rapidly accelerating.
Looking ahead, AI’s role in the world of (re)insurance will only grow. NLP will continue to streamline document processing. Machine learning models will provide increasingly accurate risk assessments. And automation will reduce the administrative burden on claims teams, allowing them to focus on high-value cases.
Crucially, AI’s success isn’t reliant on market-wide buy-in. Insurers can implement AI solutions incrementally, testing and refining them as they go. This makes AI far more amenable than blockchain, which requires an all-or-nothing commitment.
The lesson for the market here is clear. Focus your tech investments where they will deliver immediate tangible benefits. AI isn’t just another overhyped trend. It’s a proven practical tool that’s already transforming the industry. And unlike blockchain, it doesn’t require a massive leap of faith.