AI in Motion: How Free2move Is Building the Next Generation of Mobility
Free2move, by Stellantis, is a global mobility provider offering a complete and unique vehicle access ecosystem to both individual and business customers. Driven by data and technology, Free2move makes the customer experience its top priority. Clean, safe, affordable and accessible via a single app, Free2move includes free-floating car-sharing, short, medium and long-term car rental, subscription-based car-sharing and parking services.
At Free2move, artificial intelligence is not a recent add-on. It has been embedded in the company’s mobility platform for years, helping power the complex decisions required to operate at scale across millions of trips, millions of users and thousands of vehicles. Long before AI became a mainstream business talking point, Free2move built its own machine learning stack to handle data-heavy, decision-critical processes across its operations. Today, that foundation continues to evolve through a combination of in-house development, open-source technologies and ongoing testing of large language model capabilities.
Free2move first turned to machine learning out of operational necessity. In mobility, understanding demand in a dense urban environment is not something that can be managed manually. The company needed a way to learn from vast volumes of structured data and high-frequency events in real time. That led to the development of digital models of its operating cities, supported by years of granular demand data. These digital twins help Free2move predict where and when a customer is most likely to need a vehicle, enabling the business to better position cars, improve utilization and respond more dynamically to demand shifts.
That same machine learning infrastructure now supports a range of number-centric and deterministic use cases across the platform, including dynamic pricing, fleet relocation, churn prevention, risk scoring, driver authentication and other real-time decision-making processes. Rather than treating AI as a standalone feature, Free2move applies it as an operational layer across its mobility ecosystem, improving both efficiency and customer experience.
More recently, the emergence of large language models has opened a new area of opportunity: customer interaction. Free2move can now analyze and react to 100% of customer feedback comments, helping teams detect issues earlier, identify trends faster and trigger more appropriate actions. The company has tested several models to support these use cases and continues to monitor the market as solutions evolve.
One practical example is in-rental support. Free2move introduced an in-app bot to help users successfully end their rental, a process that can be interrupted by issues such as an open window, parking outside the operating zone or forgetting to turn off the engine. Before the tool was introduced, these situations accounted for roughly 10% of calls into customer service. By adding conversational support within the app, Free2move reduced total customer-service calls by about 20%, with 85% of users who interacted with the bot getting help immediately, without escalation.
The company’s experience with AI has also reinforced the importance of discipline in deployment. At Free2move’s scale, even a small system error can have major operational and financial consequences within minutes. That makes testing essential. In customer service especially, 99% accuracy is not enough when tens of thousands of users may be asking questions about pricing, terms and conditions, or how the service works. As a result, Free2move has taken a measured approach, balancing speed of innovation with the need for reliability.
Even so, AI is helping accelerate development cycles. According to Free2move, recent experimentation supported by AI coding assistants has significantly increased speed to test and deploy improvements. In one example, a new critical pricing logic moved from ideation to reliable production testing in just four weeks.
The company says it has learned three important lessons so far. First, successful AI deployment takes time, effort and training to perform specialized tasks well. Second, the strongest progress tends to come from teams and individuals who are naturally motivated to work with these technologies. Third, while many technology providers are marketing AI aggressively, far fewer truly understand how to use it effectively at scale in a live operational environment.
Looking ahead, Free2move sees the next frontier for AI in closing the gap between mass scale and true personalization in customer service. The ambition is not simply automation for automation’s sake, but a better mobility experience: fewer incidents, higher service levels and a more premium standard of support delivered to a broad consumer base.
For Free2move, AI is not just about efficiency. It is about building a mobility platform that can learn faster, respond smarter and deliver more personalized service at scale.
About Free2move
Free2move is a global mobility provider offering a complete and unique ecosystem to its individual and business customers. Driven by data and technology, Free2move makes the customer experience its top priority. Clean, safe, affordable and accessible via a single app, the offering includes free-floating car-sharing, short, medium and long-term car rental, subscription-based car-sharing and parking services. Free2move currently has more than six million customers, 450,000 rental vehicles and 500,000 parking spaces. Headquartered in Paris, the company is part of the global automotive manufacturer and mobility provider Stellantis.
For further information: https://www.free2move.com