Jack Clark, one of the co-founders of Anthropic, the business behind Claude AI, gave the analogy that AI is like the introduction of electricity. Some businesses that were steam powered put light bulbs in their businesses to make it easier for their workers to see what they were doing. But the businesses that fully grasped the implications of electricity, created their businesses to fully run on electricity and accelerated away more strongly.
It was 10 months ago that I asked, “How do you use AI in your dealership today?” In my dealer visits since then, the answer is that most dealers use AI occasionally for tasks, without structure.
There is a critical inflection point for dealerships right now. You have a powerful tool and yet you are uncertain how to best implement it. For example, ad hoc tasks done in an ad hoc way will result in duplicated effort and inconsistent output, little shared learning, data and privacy risks as well as missed opportunities. The result is little measurable business improvement and plenty of risk.
So, the question changes from “Should we use AI?” to “What outcomes should AI help us deliver?” This requires a shift from task-based thinking to outcomes-based thinking. This pathway has stepping stones; it is not one leap, and so each stepping stone might have its own set of goals – a shift in thinking and execution as your team builds confidence and capability. It’s a big internal shift to move from task or project-based AI use to strategic AI use, complete with all that that entails.
There is no doubt that AI for your dealership is a powerful tool. If you are yet to set out an AI policy for your business, perhaps right now is time to do this; create a Dealership AI Operating Model. Determine who should have paid AI subscriptions, draft some guardrails around privacy, and what information can be shared and must not be shared (you don’t want customer information in the ai learning cloud). Determine where prompts, SOPs, templates, and outputs should be stored. Who will review customer facing output, and how do departments share learning?
At its best, your AI policy should be about teamwork. This is a powerful collaborative tool if used right, there is little point in team members working in isolation and duplicating effort. We must avoid one team member seeking a solution to their issue without considering which other team members could also benefit. As I say when I run workshops with cross functional teams, the answer is not a new process created in isolation, it is the solution created in collaboration that will yield the greatest results. This is true of AI implementation.
But how do you prioritise where to is to eliminate surprises and avoid confusion after the sale has been completed.
The updated Code also reinforces an important regulatory reality for intermediaries. When offering insurance products, intermediaries are acting under each insurer’s Conduct of Financial Institutions (CoFI) licence issued by the Financial Markets Authority (FMA). This carries a clear responsibility to operate in accordance with each insurer’s Fair Conduct Programme.
In many respects, the Code aligns closely with existing regulatory obligations. Intermediaries are not simply selling a product; they are representing the insurer every time they engage with a customer.
Importantly, the Code also serves to protect intermediaries themselves. By ensuring customer needs are carefully considered and products are offered appropriately, businesses can reduce the risk of future disputes and improve overall customer outcomes.
The Code further reinforces several key customer protections, including cooling-off periods, support for vulnerable customers and access to independent dispute resolution schemes, including:
• Insurance & Financial Services Ombudsman Scheme (IFSO)
• Financial Services Complaints Limited (FSCL)
• Financial Dispute Resolution Service (FDRS)
The simplest way to view the updated Code is this: if a product genuinely meets a customer’s needs, and its benefits, features and limitations are clearly explained and understood, then the process is working as intended.
Conversely, if a product must be pushed or pressured onto a customer, it is worth questioning whether it is appropriate in the first place.
The positive message for industry participants is that this is not about adding unnecessary complexity or slowing down business. Instead, it reinforces the behaviours that good operators already demonstrate every day.
When these principles are applied consistently, customer trust grows, claims conversations become smoother and confidence in the wider industry is strengthened.
The Credit-Related Insurance Code is not about creating additional pressure; it is about establishing a consistent standard across the industry.
Nothing in the Code prevents businesses from writing strong, profitable business. Rather, it ensures that businesses are writing the right business, in the right way.
