The Bolt-On Illusion: Why AI Ambitions Collide with Legacy Core Systems

LONDON – November 10, 2026 – The global insurance industry has spent the last three years captivated by the promise of artificial intelligence. From algorithmic risk assessment to automated claims processing, the narrative has been dominated by the intelligence of the endpoint. Yet, beneath the surface of these high-tech ambitions, the digital backbone of the London Market is buckling under the pressure.

A new comprehensive study of 245 London Market professionals by insurtech firm Federato has laid bare a critical execution gap in the specialty insurance sector. While 58% of surveyed organizations currently run AI in production, a staggering 76% admit their core systems require significant modernization, transformation, or integration to actually support these advanced underwriting requirements.

The findings point to a systemic architectural friction that I have tracked for years across global supply chains and urban grids: the intelligence of a network is entirely constrained by the capacity of its underlying infrastructure. In the London Market, insurers are discovering that layering sophisticated AI over fragmented, decades-old policy administration systems does not accelerate business—it merely highlights the bottlenecks.

The Bolt-On AI Illusion

For the past decade, the prevailing strategy among many syndicates and carriers has been one of incremental digitization. Rather than ripping out legacy core systems—a process historically viewed as too costly and operationally perilous—firms opted for "bolt-on" solutions. They integrated point-solution AI tools for specific tasks, hoping to modernize their operations piecemeal.

The Federato research confirms that this era of incrementalism is ending. Only 24% of respondents now believe that minor upgrades to their existing core systems will meet future market needs.

"The research reveals a clear consensus: legacy core systems are too slow, and bolt-on AI can't deliver what it promised," said Will Ross, CEO and Co-founder of Federato. "We expect common sense to prevail: executives will realize there's no point in wasting more time and money trying to make legacy systems work, and they'll do what's necessary."

This realization comes at a critical juncture for the London Market. In March 2026, Lloyd's of London announced a strategic shift away from the grand, centralized digital transformation vision of Blueprint Two, opting instead to focus on incremental standards and data organization. By stepping back from a market-wide systemic overhaul, Lloyd's has effectively placed the burden of core modernization squarely back on the shoulders of individual managing agents, company markets, and brokers.

As one London Market chief technology officer noted on the condition of anonymity, "AI didn't solve our data silos; it made our existing data problems impossible to ignore. When you feed a high-speed algorithmic engine with fragmented, latent data from a mainframe built in the 1990s, all you get is faster access to the wrong answers."

Speed as a Survival Metric

The most alarming metric to emerge from the research is the commercial penalty of standing still. A massive 89% of firms reported losing business to competitors in the past 12 months specifically due to operational delays and a lack of speed.

In the London Market, where complex risk is syndicated across multiple parties, speed is not a luxury; it is a survival metric. The research identifies the top barriers to modernization as the very structural elements that define the market: subscription and multi-party underwriting complexity (cited by 43% of respondents) and fragmented systems across the policy lifecycle (42%). Furthermore, 38% report severe difficulties connecting their core platforms with broker, bureau, or settlement processes.

When multiple parties must agree to bind a complex risk, the slowest node in the network dictates the speed of the entire transaction. Brokers are increasingly demanding faster quotes and seamless responsiveness. In fact, faster quotes and broker responsiveness are the most frequently cited priorities for AI and core modernization (36%), outpacing the desire to improve underwriting margins (26%) or scale premium without proportional headcount growth (25%).

Firms relying on legacy systems are currently experiencing operational costs 20% to 30% higher than their digitally mature competitors. This operational drag leaves them highly vulnerable to agile, cloud-native competitors who can ingest broker data, analyze risk, and return a quote before a legacy underwriter has even finished logging into their disparate systems.

The Executive Blindspot

Perhaps the most fascinating insight from the Federato report is the profound perceptual divide between the C-suite and the operational frontline. The data reveals that C-suite executives are nearly twice as likely as operational directors (19% versus 10%) to believe that strategic underwriting changes reach the underwriting desks immediately.

The reality on the floor is starkly different. Fully 42% of professionals report that strategic underwriting changes take weeks or longer to reach day-to-day desk execution. Furthermore, reported financial losses are significantly more widespread where implementation takes longer, rising from 59% among organizations that implement changes immediately, to a staggering 100% among those that take a month or longer.

This executive blindspot is a classic symptom of disconnected infrastructure. Board members and C-suite leaders operate at the level of strategic intent. They approve a new underwriting model or a shift in risk appetite, and assume the directive is executed. However, operational directors and frontline underwriters are the ones wrestling with the digital plumbing.

Because of fragmented tools and a lack of seamless data flow, an underwriter cannot simply "execute" a new strategy. They must manually adjust parameters across multiple disconnected applications, cross-reference new guidelines against outdated policy administration systems, and coordinate via email with co-insurers. The strategy is agile; the infrastructure is paralyzed.

A Hard Core-System Reckoning

The market is beginning to respond to this existential threat, shifting capital from superficial AI experiments toward foundational overhauls. According to the research, organizations that have lost business because of operational speed are 2.4 times more likely to make a unified platform or policy administration system (PAS) modernization their top technology investment over the next three years.

This aligns with broader investment trends across the UK, where insurtech funding reached a record £2.5 billion in recent years. A significant portion of this capital is flowing toward AI-native infrastructure companies—including Federato itself, which secured a $100 million Series D funding round late last year to expand its RiskOps platform across the full policy lifecycle.

Furthermore, regulatory scrutiny is accelerating this architectural shift. As regulatory bodies increasingly demand that insurers demonstrate how and why AI-driven decisions are made, data lineage and auditability have become paramount. You cannot audit an AI decision if the underlying data is scattered across legacy spreadsheets and disconnected mainframes.

The London Market is learning a hard lesson about the nature of intelligent networks. Artificial intelligence is not a magic wand that can be waved over broken processes to make them efficient. It is a powerful engine that requires a robust, frictionless digital backbone to function. Until insurers are willing to undertake the difficult, unglamorous work of replacing their legacy core systems, the true transformative power of AI will remain firmly out of reach.