Side, a seasoned services provider, has entered into a six‑month memorandum of understanding with Denmark‑based artificial‑intelligence specialist Modl.ai. The agreement sets the stage for a joint effort to build an AI‑driven quality‑assurance (QA) platform that blends the strengths of each partner. Side, formerly known as PTW, has a long history in the services sector, while Modl.ai is a relatively new player in the gaming arena. Founded in 2018, Modl.ai secured $1.7 million in seed capital two years later and has since been experimenting with ways to apply machine learning to the testing of video games.
"The MOU essentially fuses the best of Side’s QA expertise with Modl.ai’s AI technology into a single, differentiated solution," explains Martin McBride, Side’s chief commercial officer. "One of the holy‑grail aspirations for the games industry today is to harness AI to augment QA.
No one has yet delivered a truly effective answer, and we believe we are close to that breakthrough." Earlier AI‑enabled QA tools—some of which were also developed by Modl.ai—required developers to hand over complete game builds or even source code so the testing system could run its analyses. Those approaches relied on deep integration via plugins and SDKs, creating friction and, as co‑founder Christoffer Holmgard admits, becoming "difficult over time." The new solution eliminates the need for any internal access.
"We treat the product exactly as you would hand it to a QA tester or eventually to the consumer," Holmgard says. "You load it onto a comparable device, and our system plays the game, watching the screen, simulating key inputs, and generating a comprehensive data set of everything that happens. That data is then handed to a second AI layer that flags potential bugs, escalates them to a human reviewer, and asks whether further action is required." The first phase, which Modl.ai calls "building game knowledge," involves developers uploading assets—textures, models, design documents, even screenshots—and conversing with the AI in plain language.
"All interactions are natural‑language based," Holmgard notes. "We can chat about the game while feeding the system images or design docs, allowing it to develop a contextual understanding of the title." Once the AI has a baseline comprehension, developers specify the test objectives. These can range from traditional test cases to novel scenarios that are better suited to AI‑driven exploration.
"Human expertise is still crucial at this stage," Holmgard explains, "because you need to craft prompts and language that guide the AI to perform the right actions." During a "run"—Modl.ai’s term for a testing session—the system records everything: textual logs, video snippets that serve as evidence, and all underlying telemetry. This package is then delivered to a human QA specialist, who reviews the findings and decides how to incorporate them into the production pipeline. One of the perennial challenges for automated testing is distinguishing a genuine bug from an intentional feature. Holmgard says feedback loops from human testers help the AI learn these subtleties.
"Some learning comes from the initial prompting and the pre‑built game knowledge; a large part comes from ongoing human feedback on bug reports—whether something is a defect, a feature, or an acceptable edge case. Over time the model becomes attuned to those nuances." Side envisions a future where QA personnel evolve into AI engineers, shifting the talent pool up the value chain. McBride highlights the strategic benefit: "Instead of a career path that ends at senior tester, our staff can transition into AI‑engineering roles, earn higher salaries, and open new career trajectories.
This also gives us the chance to bring jobs back to regions where they have previously vanished due to low‑cost labor pressures." According to McBride, AI will make QA work more rewarding by automating repetitive, low‑skill tasks. "Every new build demands smoke tests and countless verifications that bog down testers and limit strategic thinking. By offloading those chores to AI, our people can focus on analysis, creative problem‑solving, and higher‑order testing, which leads to a more fulfilling career at Side." Benedikte Mikkelsen, Modl.ai’s chief product officer, echoes this sentiment: "Many current tests are mind‑numbingly tedious. You don’t need a human to repeat the same flow over and over.
Freeing them from that drudgery lets them concentrate on assessing gameplay, exploring edge cases, and performing more inventive testing—that’s the real value proposition of AI." Both leaders stress that AI is not a replacement for human QA but a catalyst for upskilling. McBride notes that price pressures in certain low‑cost regions make it impossible to eliminate human testers entirely. "Providing our teams with AI toolkits enables them to extend their careers, stay relevant, and move up the value chain, which in turn helps us retain talent longer," he says. Side’s philosophy treats AI as a "sliding scale"—a spectrum where automation increases but human oversight never disappears.
"There will always be a human in the loop; that’s our ethos," McBride affirms. "The tool lets us capture the best of AI while reducing tedious work and human error, effectively raising the overall quality of the process." When asked about the types of games the system can currently handle, COO Oliver Deslandes admits the technology is best suited for simpler titles—puzzle games, card games, and other experiences with repeatable mechanics—rather than fast‑paced 3D open‑world or fighting games. "We’re focusing on titles that match the current capabilities of the AI. Clients willing to embark on this journey with us understand that the results may vary, and some will demand quick cost savings while others are more experimental." Deslandes concludes that the partnership is a long‑term experiment: "It’s a journey.
Companies that join us now for pilot projects will see how the technology matures and where it can take their QA processes in the future."