Flock Deploys AI Assistant Jay for UK Fleet Insurance Customers

Flock, a connected insurance provider specialising in commercial motor fleets, has introduced Jay, an AI-powered assistant now live in its customer portal. The tool is free for all Flock policyholders and brokers placing business with the firm.

Jay fields natural-language questions about vehicle coverage, driver conduct, claims progress, and safety ratings. Instead of static reports, the assistant queries live fleet data and delivers action-oriented insights. It also works proactively, flagging emerging risks and sending daily alerts designed to prevent incidents before they generate claims.

According to Flock, Jay was trained on more than one billion kilometres of real-world driving data. The system is configured to each fleet's individual profile rather than relying on generic industry averages.

Supported queries include vehicle-level risk warnings, month-on-month safety score trends, and checks on whether coverage remains adequate. Brokers who distribute Flock policies can access Jay across all fleets they manage, enabling them to use the tool in client reviews and renewal discussions.

Flock tested Jay with 23 fleets spanning parcel delivery, taxi, self-drive hire, and trade services. The company reports 87 per cent of pilot users returned to the assistant for follow-up queries.

One operator used Jay to discover that its most severe incidents clustered around a single road segment — an insight that would have demanded hours of manual data work, the company said.

Chief executive Ed Leon Klinger said, "For the first time in UK fleet insurance, every customer now has their own expert: someone that knows their fleet inside out, works around the clock, analyses billions of data points on their behalf, and partners with them to help their fleet run well."

The launch comes as UK commercial motor insurers increasingly adopt usage-based and telemetry-driven underwriting. Firms such as Zego and By Miles have moved toward real-time data pricing in adjacent personal and light-commercial segments.

Flock's historical edge has been integrating connected-vehicle data into both pricing and ongoing risk management, not just at the quote stage. Jay extends that approach by putting the analytical layer directly in customers' hands rather than keeping it internal.

Most AI deployments in insurance to date have targeted back-office functions like claims triage and document processing. Placing an agentic tool with end users and brokers — rather than operational staff — is a less common commercial model. It shifts the value proposition toward retention and loss-ratio improvement instead of pure efficiency gains.

Fleet insurance falls under the FCA's general insurance conduct rules, which include fair value and informed decision-making standards. An AI that surfaces coverage gaps or risk flags as part of the service could attract regulatory scrutiny if its outputs influence policy choices. Flock has not disclosed the governance or audit arrangements behind Jay's recommendations.

The next meaningful tests will be independently verifiable loss-ratio data from fleets using Jay over a full policy year, and whether broker adoption leads to measurable improvements in renewal rates.