Insurance CEOs say AI is rebalancing, not replacing, human judgement

TL;DR:

• The Sollers Consulting CEO Voices Report 2026 draws on interviews with insurance CEOs across Europe, North America and Asia-Pacific, and finds AI shifting from experimentation to day-to-day operational use.
• The central editorial line — that AI should make human judgement more important alongside automation, not redundant — is at odds with the cost-cutting rhetoric heard from US enterprise vendors this week, including Cisco’s 4,000-job AI-focused restructuring.
• The report places the trend in the context of 25 years of insurance technology change and identifies risk management, workforce transformation and model governance as the three pressure points UK carriers should watch.

The “AI alongside human judgement” framing is a deliberate response to industry fatigue with the “AI will replace underwriters” coverage of 2024-2025. Sollers president Marcin Pluta is explicit: “Our interviewees reject the notion that AI inevitably destroys jobs.” Instead, executives describe gradual change in the skills profile — more data literacy, more analytical capacity, less rekeying and triage. That maps closely to the same workforce-transformation framing in this week’s Zopa-led Jobs2030 coalition launch in UK financial services.

Where the operational changes are landing

The report identifies four AI use clusters now in production: automated ingestion of broker submissions and claims documents; AI-assisted triage and fraud detection; virtual assistants in customer and broker service; and information-synthesis tools supporting underwriters and claims handlers. None are pilots — Sollers describes the shift from experimentation to operational use as the year-over-year change. Sollers president Michał Trochimczuk frames cost control as the strategic driver: “investing in automation, simplified standard architectures, modern rating systems, and intelligent risk selection”.

UK angle: where the regulatory perimeter is moving

For UK insurance leaders, the practical question is how AI governance maps onto Solvency II, the FCA’s Consumer Duty, and the Pensions Regulator’s new AI guidance (covered separately this week). The Sollers report flags model governance, bias and explainability as the operational concerns boards are now actively managing — the same concerns Logicalis’s CIO survey this week showed UK CIOs only 31% confident on. Regulators have signalled increasing scrutiny of pricing, underwriting and claims decisions specifically; AI deployments touching those areas now carry conduct and prudential risk in addition to operational risk.

Looking forward

Three questions follow for UK insurers. First, whether the “AI is augmenting, not replacing” framing survives the next downturn — historically, technology that allows headcount reduction tends to be deployed that way under pressure. Second, whether UK-specific frameworks emerge for AI in claims adjudication, particularly for vulnerable customer journeys. Third, how external-data use in property and specialty lines — the area Sollers identifies as undergoing the most change — interacts with the UK’s data-protection framework as the AI Act Omnibus deadlines reshape EU adjacent obligations.