Generic artificial intelligence platforms such as ChatGPT, Claude, Gemini and Copilot are rapidly becoming part of day-to-day operations inside UK dealer groups, but retailers are being warned against treating them as simple plug-and-play solutions, writes Tom Seymour.
From drafting emails and summarising meetings to lead nurturing and operational workflows, dealerships are increasingly experimenting with large language models (LLMs) as the technology evolves at pace.
Yet alongside the productivity gains come growing questions around security, governance, data ownership and whether generic artificial intelligence (AI) tools are sufficient to cope with the complexities of automotive retail.
OpenAI and Anthropic, the companies behind ChatGPT and Claude respectively, are iterating their models at a speed few enterprise software providers can match.
Industry analysts note that OpenAI’s major model updates are now arriving roughly every three-to-four months, compared with previous six-to-nine-month cycles, while businesses are simultaneously trying to build long-term operational strategies around technology that is changing continuously.
That rapid evolution is also fuelling extraordinary investor valuations across the sector.
OpenAI’s $852 billion (£635bn) valuation has climbed sharply as businesses globally increase adoption of generative AI systems, while Anthropic has also secured enormous backing amid expectations that enterprise AI spending will continue to accelerate.
For dealers trying to decide whether to commit to one provider, build automotive-specific workflows or rely on third-party suppliers, the pace of change creates both opportunity and uncertainty.
According to the Department for Science, Innovation and Technology’s AI Adoption Research from February 2026, businesses already using AI are overwhelmingly relying on natural language processing and text generation systems, with 85% of users deploying those capabilities.
By contrast, only 7% were using agentic AI systems, which go beyond answering questions and can perform complex multi-step actions from system to system.
Half of businesses using AI were deploying more than one AI technology.
Inside automotive retail, many dealer groups are now moving beyond experimentation and starting to ask where AI genuinely adds operational value.
OpenAI and Anthropic may dominate headlines, but automotive suppliers argue that the real challenge for retailers is understanding the difference between general-purpose productivity tools and AI systems built specifically for automotive workflows.
Three different AI conversations
James Leese, UK managing director at automotive AI specialist Impel, says dealerships are currently navigating “three completely different AI conversations happening simultaneously” and warns that many businesses are blurring them together.
Leese says: “If I go back two or three years, having conversations about AI and bringing AI into businesses left most people scratching their heads.
“There was concern around security and whether AI was going to take over jobs. Now businesses are starting to engage with it and experiment with it, which is very healthy.”
However, he says many dealer groups are failing to separate consumer AI tools from operational AI platforms and enterprise automotive systems.
Identifying operational issues
Leese says problems often begin when businesses start with the technology rather than identifying the operational issue they are trying to solve.
“What problem am I trying to solve? If it’s simply nurturing leads outside business hours because 30% or 40% of enquiries arrive when the dealership is closed, AI can already help with that very effectively.
“But if businesses jump straight to saying: ‘I want AI to run my business,’ then that’s the wrong mindset.”
He compares AI deployment with onboarding a new employee into a dealership. “You wouldn’t hire a salesperson on day one, give them no training and say ‘off you go’,” says Leese.
“The same applies to AI. If the system can access lead information, it can have intelligent conversations around leads.
“If it can access stock information, it can answer questions about vehicles. But if it cannot access finance information, then you absolutely do not want it attempting finance conversations because that creates compliance risks.”
Creating an AI brain
Leese says larger dealer groups are increasingly focusing on how AI systems access and structure dealership data rather than simply purchasing standalone tools.
“There’s now a growing view around creating an ‘AI brain’ inside the business,” he says. “If you just bolt multiple AI products onto different areas of the operation, the consumer journey can become very disjointed.”
He says enterprise-scale dealer groups often already operate across multiple dealer management systems (DMSs) and customer relationship management (CRM) platforms, making integration strategy increasingly important.
“Trying to force everything through one platform isn’t always realistic,” he says. “What matters is whether the data can flow properly between systems and whether the customer experience remains consistent.”
Leese also believes OEMs are becoming more actively involved in shaping how AI develops inside their retail networks.
He describes conversations with manufacturers around building centralised “knowledge banks” that could feed approved information into AI systems used by retailers.
“It means AI systems can access accurate car manufacturer information without every individual retailer having to build that themselves.”
Generic vs industry-specific AI
That distinction between generic AI and automotive-focused systems is becoming a growing battleground across the sector.
Peter Appleby, head of data science and analytics at Autotrader UK, says generic AI tools can support productivity improvements, but warns they often struggle with the operational complexity of automotive retail.
“AI is already helping retailers save time and remove friction from core operational tasks,” he says. “However, general-purpose AI tools can only go so far.
“Automotive retail has some very specific requirements and simply bolting generic tech onto legacy systems often leads to disjointed workflows.”
Appleby says specialist automotive systems gain an advantage through proprietary datasets and tighter safeguards.
He believes that when AI is used in areas where accuracy, consistency and financially significant decisions matter, the quality of the data behind it and the safeguards around it become critical.
He points to Autotrader’s Co-Driver platform, which is built using the company’s proprietary vehicle, valuation and consumer audience data.
“Generic tech models simply cannot replicate that out of the box,” he says.
Appleby says the business has focused heavily on monitoring and safeguards designed to reduce hallucinations and inaccurate outputs. AI hallucinations can happen where a result generates false, misleading or entirely fabricated information that sounds confident, highly structured and perfectly logical.
Those who have used ChatGPT or Claude may have experienced this and it is up to humans using these systems to catch these mistakes.
According to Appleby, Autotrader’s Co-Driver tools are now being used by around 11,000 retailers, representing roughly 85% of its customer base, with an estimated 200,000 hours of manual work saved collectively.
However, despite promoting specialist solutions, Appleby does not see the future as automotive businesses competing directly against large AI providers.
He says: “The major AI platforms will continue to innovate quickly. The opportunity for automotive businesses is not to compete with them, but to build on their foundational capabilities to create auto-specific solutions.”
He points to how automotive platforms are adapting to conversational AI search.
He adds: “As consumer search habits shift towards conversational AI, retailers still need their stock to remain visible and discoverable.
“Ultimately, the key is ensuring AI is developed responsibly by trusted partners who can integrate it properly into existing workflows.”
Security risks dealers may overlook
Alongside operational concerns, dealer groups may be underestimating the risks associated with widespread use of generic AI tools.
Lisa Ventura, founder of Cyber Security Unity, told AM many automotive businesses have embraced AI tools without fully understanding the implications around data handling and governance.
Ventura says: “The biggest concern is data leakage. Customer data, financial information, vehicle stock details, registration numbers, internal processes and even pricing strategies could potentially be entered into these systems.”
She warns that dealership employees using consumer-grade or standard business AI subscriptions may unknowingly expose commercially sensitive information.
Ventura says many employees are primarily focused on improving productivity rather than scrutinising provider terms and data-handling policies.
She also highlights growing concern around “shadow AI”, where staff adopt AI tools independently without approval or oversight from IT departments. That creates a lack of visibility over what information is leaving the business. “It also makes it very difficult to enforce internal policies around customer data handling.”
Ventura warns this could create UK GDPR compliance risks because dealer groups remain responsible for how customer data is processed, even when third-party AI systems are involved.
Ventura also cautions dealers against assuming automotive-specific AI suppliers automatically provide stronger security. “Sector-specific positioning should not automatically be viewed as a security advantage.”
Instead, she says dealer groups should ask detailed questions around data storage, model training, access controls and certifications such as ISO 27001 or Cyber Essentials Plus.
“A reputable provider should be able to answer those questions clearly and in writing,” she adds.“Vague responses should be treated as a red flag.”
Subscriptions, tokens and costs
One factor making AI adoption more confusing for dealers is the increasingly complex commercial structure behind modern AI platforms.
Unlike traditional automotive software subscriptions, many AI systems combine user licences with token-based usage pricing.
AI tokens are the basic units of data used by LLMs to process language. Rather than analysing entire words, systems break text into smaller fragments which are converted into numerical data for processing.
Every prompt submitted and every response generated consumes tokens, meaning costs can scale rapidly depending on how extensively businesses integrate AI into workflows.
For many dealerships, entry-level adoption starts with standard business subscriptions for tools such as ChatGPT Team, Claude Team, Microsoft Copilot or Gemini for Workspace, typically charged at a per-employee rate for each month.
However, more advanced integrations often move onto API-based pricing structures, where businesses pay according to usage volumes rather than fixed user licences.
Anthropic, for example, offers prepaid enterprise credits and token-based API billing alongside standard subscription plans.
Businesses deploying AI heavily into lead management, customer communications or automated workflows may consume a large number of tokens each month, meaning monitoring usage becomes increasingly important.
Leese says this is another reason why retailers need a clear strategy before scaling deployments.
“The challenge is not just introducing AI,” he says.
“It’s understanding where experimentation ends and operational dependency begins.”
For now, the consensus across the sector appears to be that generic AI tools are likely to remain part of dealership operations, particularly for productivity tasks and experimentation.
But, as retailers move towards embedding AI deeper into customer journeys, operational systems and data flows, pressure is increasing to build more structured, secure and automotive-specific approaches.
As Leese puts it: “Businesses should stop starting with AI as the solution and instead ask what challenge they are trying to solve.”
This feature first appeared in the 2026 AM Dealer Technology Guide which brings together leading technology providers to showcase the latest innovations available to help dealerships improve performance across both customer-facing and back-office functions. Read the report HERE