Tools & Platforms
Will AI transform the way used cars are bought and sold?
03 July 2025
This year has seen a surge in artificial intelligence (AI) advances. But what impact has this technology made on the used-car retail industry, and what is yet to come? Autovista24 journalist Tom Hooker takes a deep dive into the subject.
Through the likes of ChatGPT, Google Gemini and Microsoft Copilot, AI has transformed the way we work. Forbes reported that the technology will reach a market revenue of $1.33 billion (€1.18 billion) by 2030. Meanwhile, 64% of businesses believe that artificial intelligence will help increase their overall productivity.
Within the automotive sector, AI is already embedded in manufacturing and quality control, such as BMW’s ‘Factory Genius’ assistant. It is also being used to improve connected car experiences. Volvo Cars is using AI to enhance advanced driver-assistance systems (ADAS).
How would the technology work in the world of used-car retail? It could give customers a more personal and efficient experience. But how does this translate into realistic sales and revenue growth for dealerships?
AI and disruption
Answering this question means stepping back to look at the AI industry and the anticipated changes just around the corner.
‘Now we see what we believe to be also a highly disruptive change coming up with artificial intelligence,’ stated McKinsey & Company partner Peter Cholewinski at the Used Vehicle Retail Summit.
‘The topic is not new. AI has been around for many years. However, with the introduction of ChatGPT, this has arrived in our daily lives and in the lives of companies. The speed of progress is just amazing,’ he added.
ChatGPT is an example of a generative large language learning model (LLM). This means it can create content such as text and images in response to a person’s prompt or request.
To do this, it relies on using machine frameworks known as deep learning models. These algorithms simulate the human brain’s learning and decision-making processes.
Cholewinski showed the growing number of LLM releases. In 2024, 122 new models entered the market. This was up from the 109 LLMs launched in 2023 and a significant increase from 29 releases in 2022.
From left to right: Peter Cholewinski, McKinsey & Company partner. Dr Lisa Schrewentigges, McKinsey & Company project manager.
‘In 2025, you have many models out there, and those models are becoming smarter. We are now not talking about large language models, but about reasoning models. Additional tools are also coming out, like deep research. The machine can go on its own onto the internet and figure a lot of information out by itself,’ Cholewinski explained.
Agentic AI can capture value
While generative AI LLMs depend on users’ prompts and requests, agentic AI LLM models are designed to autonomously make decisions and act, with the ability to pursue complex goals with limited supervision, IBM wrote. This combines the flexibility of LLMs with the accuracy of traditional programming.
‘This year, everybody is talking about agentic AI. When you take those models with reasoning capabilities, they can plan and think about what they need to do to achieve a goal. You can also have several of them working together, exchanging basic information, reviewing each other, and trying to solve a problem on their own.
‘So, it is not only about one chatbot that you talk to, but end-to-end processes and how several agents can achieve something useful and valuable.
‘Agentic machines can tap into different workflow steps and coordinate across those workers. This means we have more automation possibilities across workflows. This is where most of the value will be captured, especially as they become smaller,’ commented Cholewinski.
The first fully autonomous agentic LLM model, Manus, was released in March 2025, as written by Forbes.
AI transformation troubles
‘Everybody is trying it out, but only a very small number can say we invested something, and we actually captured something. This is because it is very difficult,’ said Cholewinski.
‘You need to have the technology, but you also need to have the right talent to understand how to use that technology and an operating model that will drive the change management to scale and adopt this technology,’ he added.
Cholewinski showed that 88% of companies attempt a digital and AI transformation. However, just 25% meaningfully progress in their digital and AI transformation. Furthermore, just 10% of enterprises have AI at scale, and under 5% of scale use-cases deployed are active across full workflows.
‘In the cases where we are seeing value being captured, they are thinking about several use cases together and in an agentic fashion,’ he highlighted.
AI assists dealership leads
So, what real-world use cases are already being implemented in the automotive retail sector, and what impact is this having?
One example is a generative AI-based tool that can tailor and personalise messages for customers and online leads. The unnamed product was built for one of the largest German dealer groups. This means covering 200 different dealerships and a database of over 500,000 existing customers from vehicle sales.
‘What they struggled with is looking into the lead management and how to have a very structured approach in contacting existing customers in a very fast way, which is also very tailored,’ explained McKinsey & Company project manager Dr Lisa Schrewentigges.
In her presentation, she showed that the dealer group previously spent around five to 10 minutes on every customer outreach. They also struggled with how to respond to incoming website leads and how to personalise this interaction.
Fast development times
‘What we have done together with them is, within six weeks, develop a generative AI tool, which allowed them to identify the most promising leads. Secondly, tailor the messages towards those leads and be fast in answering those leads,’ she outlined.
‘With generative AI and agentic AI, you can implement those kinds of solutions very fast because you do not need to train the AI anymore. These models are so powerful that you can actually use them off the shelf,’ noted Cholewinski.
‘This is also where the potential lies. You can think about your end-to-end processes, where there is a lot of manual work that you could improve. Then, think about the several use cases that make sense to improve productivity or sales with this technology,’ he added.
The sales agent journey
Schrewentigges walked through the typical sales agent journey. This starts with selecting a customer and thinking about which promotion to send. Then, interacting with the customer, and in the end, moving this customer towards a decision.
‘Where we helped here was bringing together the customer information that they already have on the system, matching it with third-party data and different website data,’ she said.
‘Then you have a full, enriched customer profile, identifying the most promising leads and personalising communications with a specific customer, which helps the sales agent convert them to a sale,’ Schrewentigges said.
A dashboard then enables the sales agent to see a full customer overview. It can prioritise the customer based on a lead score and suggest specific email campaigns. The dashboard also displays different customer groups, such as existing customers, website leads, and follow-ups.
She then showed the typical outcome of this personalised messaging. Various data points can be used by the generative AI to create an individualised email to the customer.
‘They were able to not only send out emails, but also very personalised phone calls based on the information that we put together. This, in the end, led to much faster reply times from website leads, because we had a very standardised approach in answering typical emails, but also it led to much more personalised communication,’ Schrewentigges said.
An instant impact?
‘We had a lot of impact regarding the speed of answers, personalised communication, but also in the end, this will ultimately sell cars much faster,’ she stated.
The dealer group recorded an increase of more than 20% in conversion rates. Each sales representative recorded an additional 15 to 25 vehicle sales annually on average.
This was made possible through a 70% to 80% efficiency gain, which meant more time to sell cars. Furthermore, 10 to 15 times more customers were approached with relevant sales campaigns. However, there were still significant challenges and concerns for the tool to overcome.
‘You always need to drive a balance between not using too much information because once you go into too many details that the AI might know, it becomes very creepy,’ commented Cholewinski.
Additionally, as AI becomes more powerful, could this put jobs in dealerships at risk in the future?
‘Even though generative AI solutions will help with emails, there will always be a personalised component in contacting the dealership, having a phone call, and visiting the car,’ said Schrewentigges.
‘I think it will, in a certain part, probably affect how vehicles will be sold, but we always need this component. People come to the dealership and want to see and feel a vehicle,’ she explained.
Virtual assistants for retailers
Elsewhere, Novaco AI provides virtual assistants that can be used on automotive retailer websites. By connecting to their data, the assistants are designed to improve dealership efficiency, automate conversations, and optimise customer interactions.
‘It is connected to inventory, virtual planning, digital work orders, but also your lead management system,’ outlined Novaco AI CEO Maarten Bekkers.
The assistant started with Google AI in 2019. After LLMs were released, the tool began utilising them. It is now beginning to use agentic AI models and is bringing its assistant to WhatsApp.
The company also provides a virtual assistant for dealership employees to increase their efficiency and find information quickly. The AI companion is also connected to pricing information.
‘So, if somebody calls and asks, “what would it cost to replace my clutch for the car with that number plate?”, you just fill in the question to the companion and it will generate the answer within a few seconds. ‘It is a real virtual employee that works for you,’ said Bekkers.
The assistants can also help dealerships with common queries, freeing up time for employees.
‘Complaint number one at dealerships is that the phone keeps on ringing with the same questions every day. The majority of people who book a service call the dealer. It is the most expensive resource of the dealer is actually booking the service, it is crazy,’ he commented.
‘So, you should turn it around. If people really want to call, they can still call. But in the near future, a virtual assistant will be on the phone, having the same conversation as a human and making a booking,’ Bekkers added.
AI’s organisational prowess
‘AI has been instrumental for our success,’ said Lizy’s chief of staff, Nicolas Daive, as he began his presentation. The company is an online B2B car leasing platform offering used vehicles to companies.
‘Used cars are more operationally complex and messy than new cars. Despite that, because you have lower asset value, lower leasing prices and longer holding periods, you can be extremely efficient. With AI, we were able to transform this messy product into a very simple operation,’ highlighted Daive.
‘To make sure we have the best possible offering, we source vehicles all over Europe, across more than 100 suppliers. This means that we have more than 100 data formats, data types, processes, and ways of working.
‘In the past, working with this number of suppliers would have meant you needed four or five full-time employees due to the complexity it brings. With AI, we were able to do this with half a full-time employee,’ he commented.
Daive explained the process of buying cars from a supplier, with a PDF containing data. An employee then forwards the PDF to their AI agent with a few instructions. This includes scheduling a pickup time for the vehicles and pre-pricing them.
‘All that is done from the click of a button. In the past, we probably would have had a full-time employee that is doing a lot of copy and pasting, getting the right data into the right fields, and talking to a lot of departments,’ he noted.
‘Automation is nothing new. Commission is something we have been doing for almost four decades. What is new is that AI allows us to automate chaos. It can take unstructured data, structure it, then send it to the right places,’ concluded Daive.
Exploring the critical role of technology in advancing Harvard’s mission and the potential of generative AI to reshape the academic and operational landscape were the key topics discussed during University’s 12th annual IT Summit. Hosted by the CIO Council, the June 11 event attracted more than 1,000 Harvard IT professionals.
“Technology underpins every aspect of Harvard,” said Klara Jelinkova, vice president and University chief information officer, who opened the event by praising IT staff for their impact across the University.
That sentiment was echoed by keynote speaker Michael D. Smith, the John H. Finley Jr. Professor of Engineering and Applied Sciences and Harvard University Distinguished Service Professor, who described “people, physical spaces, and digital technologies” as three of the core pillars supporting Harvard’s programs.
In his address, “You, Me, and ChatGPT: Lessons and Predictions,” Smith explored the balance between the challenges and the opportunities of using generative AI tools. He pointed to an “explainability problem” in generative AI tools and how they can produce responses that sound convincing but lack transparent reasoning: “Is this answer correct, or does it just look good?” Smith also highlighted the challenges of user frustration due to bad prompts, “hallucinations,” and the risk of overreliance on AI for critical thinking, given its “eagerness” to answer questions.
In showcasing innovative coursework from students, Smith highlighted the transformative potential of “tutorbots,” or AI tools trained on course content that can offer students instant, around-the-clock assistance. AI is here to stay, Smith noted, so educators must prepare students for this future by ensuring they become sophisticated, effective users of the technology.
Asked by Jelinkova how IT staff can help students and faculty, Smith urged the audience to identify early adopters of new technologies to “understand better what it is they are trying to do” and support them through the “pain” of learning a new tool. Understanding these uses and fostering collaboration can accelerate adoption and “eventually propagate to the rest of the institution.”
The spirit of innovation and IT’s central role at Harvard continued throughout the day’s programming, which was organized into four pillars:
- Teaching, Learning, and Research Technology included sessions where instructors shared how they are currently experimenting with generative AI, from the Division of Continuing Education’s “Bot Club,” where instructors collaborate on AI-enhanced pedagogy, to the deployment of custom GPTs and chatbots at Harvard Business School.
- Innovation and the Future of Services included sessions onAI video experimentation, robotic process automation, ethical implementation of AI, and a showcase of the University’s latest AI Sandbox features.
- Infrastructure, Applications, and Operations featured a deep dive on the extraordinary effort to bring the new David Rubenstein Treehouse conference center to life, including testing new systems in a physical “sandbox” environment and deploying thousands of feet of network cabling.
- And the Skills, Competencies, and Strategies breakout sessions reflected on the evolving skillsets required by modern IT — from automation design to vendor management — and explored strategies for sustaining high-functioning, collaborative teams, including workforce agility and continuous learning.
Amid the excitement around innovation, the summit also explored the environmental impact of emerging technologies. In a session focused on Harvard’s leadership in IT sustainability — as part of its broader Sustainability Action Plan — presenters explored how even small individual actions, like crafting more effective prompts, can meaningfully reduce the processing demands of AI systems. As one panelist noted, “Harvard has embraced AI, and with that comes the responsibility to understand and thoughtfully assess its impact.”
Source link
Some tennis players are not happy with Wimbledon’s new AI line judges, as reported by The Telegraph.
This is the first year the prestigious tennis tournament, which is still ongoing, replaced human line judges, who determine if a ball is in or out, with an electronic line calling system (ELC).
Numerous players criticized the AI technology, mostly for making incorrect calls, leading to them losing points. Notably, British tennis star Emma Raducanu called out the technology for missing a ball that her opponent hit out, but instead had to be played as if it were in. On a television replay, the ball indeed looked out, the Telegraph reported.
Jack Draper, the British No. 1, also said he felt some line calls were wrong, saying he did not think the AI technology was “100 percent accurate.”
Player Ben Shelton had to speed up his match after being told that the new AI line system was about to stop working because of the dimming sunlight. Elsewhere, players said they couldn’t hear the new automated speaker system, with one deaf player saying that without the human hand signals from the line judges, she was unable to tell when she won a point or not.
The technology also met a blip at a key point during a match this weekend between British player Sonay Kartal and the Russian Anastasia Pavlyuchenkova, where a ball went out, but the technology failed to make the call. The umpire had to step in to stop the rally and told the players to replay the point because the ELC failed to track the point. Wimbledon later apologized, saying it was a “human error,” and that the technology was accidentally shut off during the match. It also adjusted the technology so that, ideally, the mistake could not be repeated.
Debbie Jevans, chair of the All England Club, the organization that hosts Wimbledon, hit back at Raducanu and Draper, saying, “When we did have linesmen, we were constantly asked why we didn’t have electronic line calling because it’s more accurate than the rest of the tour.”
We’ve reached out to Wimbledon for comment.
This is not the first time the AI technology has come under fire as tennis tournaments continue to either partially or fully adopt automated systems. Alexander Zverev, a German player, called out the same automated line judging technology back in April, posting a picture to Instagram showing where a ball called in was very much out.
The critiques reveal the friction in completely replacing humans with AI, making the case for why a human-AI balance is perhaps necessary as more organizations adopt such technology. Just recently, the company Klarna said it was looking to hire human workers after previously making a push for automated jobs.
The Raspberry Pi Foundation’s analysis presents coding as not merely a vocational skill but a fundamental literacy that develops critical thinking, problem-solving abilities, and technological agency — competencies argued to be increasingly vital as AI systems permeate all aspects of society. The foundation emphasizes that while AI may automate certain technical tasks, human oversight remains essential for ensuring the safety, ethics, and contextual relevance of computer-generated solutions.
For educators, parents, and policymakers, this report provides timely insights into preparing younger generations for an AI-integrated future.
Image Credit: Raspberry Pi Foundation
-
Funding & Business7 days agoKayak and Expedia race to build AI travel agents that turn social posts into itineraries
-
Jobs & Careers7 days ago
Mumbai-based Perplexity Alternative Has 60k+ Users Without Funding
-
Mergers & Acquisitions7 days ago
Donald Trump suggests US government review subsidies to Elon Musk’s companies
-
Funding & Business6 days ago
Rethinking Venture Capital’s Talent Pipeline
-
Jobs & Careers6 days ago
Why Agentic AI Isn’t Pure Hype (And What Skeptics Aren’t Seeing Yet)
-
Funding & Business4 days ago
Sakana AI’s TreeQuest: Deploy multi-model teams that outperform individual LLMs by 30%
-
Funding & Business7 days agoFrom chatbots to collaborators: How AI agents are reshaping enterprise work
-
Jobs & Careers6 days ago
Astrophel Aerospace Raises ₹6.84 Crore to Build Reusable Launch Vehicle
-
Funding & Business6 days ago
Europe’s Most Ambitious Startups Aren’t Becoming Global; They’re Starting That Way
-
Jobs & Careers4 days ago
Ilya Sutskever Takes Over as CEO of Safe Superintelligence After Daniel Gross’s Exit
