Dassault Systèmes plans to acquire ArisGlobal for up to $2 billion.

      Dassault Systèmes ranks among the largest software companies in Europe, and this year has been quite challenging for it. Its stock has declined by about 25% this year. The chairman was ousted in February, and investors are concerned that a new wave of AI technology could undermine its core business.

      In response to these challenges, Dassault announced on Thursday that it plans to spend up to $2 billion on a relatively unknown company.

      The French company has reached an agreement to acquire ArisGlobal, a provider of drug safety and regulatory software, for approximately $1.8 billion in cash, plus up to $200 million more contingent on future AI revenue. The seller is Nordic Capital, a Swedish private equity firm that has owned ArisGlobal since 2019, and the entire amount will come from Dassault’s cash reserves.

      What Dassault is acquiring

      ArisGlobal may not be a familiar name, but it plays a crucial role in the pharmaceutical industry. Founded in 1989 and headquartered in Waltham, Massachusetts, it caters to over 200 clients, including half of the largest 50 drug manufacturers globally, as well as biotechs and government health regulators. Its software manages more than 12 million patient safety reports annually, with projected revenue of around $175 million by 2026.

      The appeal for Dassault lies in the AI capabilities of ArisGlobal. The NavaX system utilizes generative AI for the labor-intensive processes of drug safety monitoring and regulatory submissions, claiming to reduce that workload by over 30%. The market for compliance software of this type could reach $7.5 billion by 2030.

      Dassault isn’t the only company making acquisitions; established software firms like SAP and Adobe are also opting to buy into AI rather than develop it internally.

      The virtual twin investment

      Dassault specializes in creating “virtual twins,” which are intricate digital replicas of actual machines, factories, and increasingly, human bodies. These digital twins assist 390,000 clients in designing cars, aircraft, and pharmaceuticals. CEO Pascal Daloz aims to integrate ArisGlobal's real-world patient data with these virtual models. He envisions creating the first platform to "combine real-world evidence with virtual evidence."

      “Life Sciences is on the brink of a new innovation era, spurred by AI and the transition to precision medicine,” Daloz stated.

      ArisGlobal CEO Aman Wasan expressed a more direct view, saying, “The future of Life Sciences will not be defined by superior software, but by interconnected intelligence.”

      A deal under pressure

      This enthusiasm is overshadowed by significant challenges. Dassault has faced scrutiny in 2026 as investors worry that AI could disrupt the software sector. One specific concern is the emergence of “world models,” which enable machines to understand the physical world without relying on a meticulously crafted twin. The stock saw a sharp decline, resulting in the abrupt departure of the chairman in February. This acquisition marks Dassault’s first significant move since then.

      Complicating matters, the division that Dassault is choosing to invest in is also its weakest. According to Bloomberg, life sciences revenue dropped 4% in Q2 and 12% in Q1, negatively impacting overall results. Its prior significant investment in the healthcare sector, the 2019 acquisition of clinical trials company Medidata, has not met expectations. Now, it is once again betting on this sector, hoping that AI can transform its underperforming segment into a growth driver.

      For the time being, investors have reacted positively. Dassault’s shares increased by about 5% on Thursday. The company also reported solid quarterly performance, with revenue rising 4% and cloud services growing by 14%. The transaction is expected to close in the second half of 2026, pending regulatory approval. However, whether this move can assuage broader concerns that AI may also threaten the creators of digital twins remains a more complicated issue.