AI underwriting

Category: Insurtech · Reviewed by Matt Bartlett, Director · Founder · Last reviewed 2026-06-05

AI underwriting

AI underwriting is the application of machine learning, large language models, computer vision and other artificial intelligence techniques to insurance underwriting decisions. Unlike traditional algorithmic underwriting based on declared variables and statistical rating, AI underwriting may extract underwriting-relevant features from unstructured data (text, images, satellite imagery, IoT streams) and apply non-linear learned models. It is the subject of specific FCA/PRA supervisory expectations.

Category: Insurtech Also known as: Artificial intelligence underwriting, Machine learning underwriting Established: Concept early 2010s; mass deployment since c. 2018; LLM applications since c. 2023 Related concepts: Algorithmic underwriting, Insurtech, Lemonade Insurance

Definition

AI underwriting techniques include: supervised machine learning (gradient boosting, random forests, neural networks) for risk classification and pricing; large language models (LLMs) for parsing proposal forms, broker submissions, and underwriting referrals; computer vision for property and vehicle damage assessment; satellite imagery for property risk assessment and natural hazard exposure; and reinforcement learning for portfolio steering. The boundary between “advanced algorithmic underwriting” and “AI underwriting” is not sharp — the term AI is typically applied to non-linear learned models and to language- or vision-based feature extraction.

Legal / Regulatory basis

The FCA and PRA joint Discussion Paper DP5/22 Artificial Intelligence and Machine Learning (October 2022) and Feedback Statement FS2/23 (October 2023) set out the UK supervisory framework. Key requirements: governance and accountability for AI decisions (SM&CR); data quality and integrity; model risk management; transparency and explainability; consumer outcomes; and bias and fairness. UK GDPR Article 22 (automated individual decision-making) and the broader transparency requirements (Articles 13/14) apply.

The EU AI Act (Regulation (EU) 2024/1689, entry into force August 2024 with phased application) classifies certain AI uses in insurance as “high-risk” with consequent governance, documentation and conformity obligations. The UK has not adopted the EU AI Act framework and has indicated a more principles-based approach via existing regulators.

How it works in practice

A modern Lloyd’s syndicate may use AI to: triage incoming broker submissions; extract policy schedule data and exposure information from unstructured PDF submissions; price the placement using a hybrid GLM/gradient boosting model; recommend additional terms or exclusions; and identify portfolio aggregation. LLM-based submission parsing has reduced underwriting cycle times in some markets from days to minutes for routine business.

Lemonade Insurance has been a prominent AI-led insurer, using chatbots and AI claims triage. Specialist insurtechs such as Cytora and other underwriting AI vendors provide enterprise-grade AI underwriting platforms to traditional insurers.

Common variations

Decision-support AI (supporting human underwriter decisions); decision-making AI (replacing human decisions in defined scope); generative AI (using LLMs to draft policy wording or correspondence); and embedded AI (AI as a component within traditional underwriting workflow).

Example

A specialty professional indemnity insurer using an LLM-based submission triage system to extract risk variables from a broker presentation, supplemented by a gradient boosting model for pricing on a defined retention layer, with human underwriter sign-off above a pre-set threshold and full audit trail to satisfy SM&CR accountability.

See also

References

  1. FCA / PRA Joint DP5/22 AI and Machine Learning (October 2022) — https://www.fca.org.uk
  2. FCA / PRA Feedback Statement FS2/23 (October 2023)
  3. Regulation (EU) 2024/1689 (EU AI Act)
  4. UK GDPR Articles 13, 14, 22
  5. SS1/23 Model risk management principles for banks (PRA, May 2023) — applied by analogy to AI in insurance

This entry is part of the Apex Insurance Wiki. Last reviewed by Matt Bartlett on 2026-06-05. Next review: 2026-12-05.

Apex Insurance Brokers Limited. Authorised and regulated by the Financial Conduct Authority, FRN 724952. Registered in England and Wales, Companies House 07014570. This entry provides general information about UK insurance concepts and is not regulated advice. Consult your insurance broker on your specific position.

Talk to a specialist broker

Apex Insurance Brokers serves UK professional services firms and commercial businesses. Call 0117 325 0027, email hello@apexinsurancebrokers.co.uk, or request a quotation.

Get a quote
Our service promise. We acknowledge every quote request the same working day. For straightforward risks, indicative terms typically follow within five working days. Complex risks — higher-risk buildings, cladding, mid-term proposals requiring fresh underwriting — may take longer; we’ll send you a progress note by the end of the fifth working day in those cases.
★ 4.0 on Trustpilot (verified)|Listed on the ARB PI broker list|FCA FRN 724952