Side, a seasoned services provider formerly known as PTW, has entered into a six‑month memorandum of understanding (MOU) with Denmark‑based artificial‑intelligence specialist Modl.ai. The agreement binds the two companies to co‑develop an AI‑enhanced quality‑assurance (QA) platform that aims to reshape how video‑game testing is performed. Modl.ai, a relatively new player in the gaming sector, was founded in 2018 and secured $1.7 million in seed financing two years later. Side, by contrast, has been operating in the services arena for many years, offering a deep bench of QA expertise and a global delivery network.
By joining forces, the two firms hope to combine Side’s extensive testing know‑how with Modl.ai’s cutting‑edge machine‑learning algorithms to create what Martin McBride, Side’s chief commercial officer, describes as “the crown‑jewel, holy‑grail experience for the games industry.” Historically, AI‑driven QA solutions have required developers to hand over full game builds—or even source code—to the testing vendor. Modl.ai’s earlier offering relied on an integration model that used a plugin and software development kit (SDK), forcing developers to expose internal assets. This approach created significant friction, as co‑founder Christoffer Holmgard explains, and eventually proved unsustainable.
The new system eliminates the need for any internal code access. Holmgard likens the process to handing the product to a traditional QA tester or, ultimately, to a consumer. The game is installed on a device that mirrors the target platform, and the AI watches the screen, simulates player inputs, records every visual and telemetry event, and then passes the resulting data set to a second AI layer that flags potential bugs.
Those flagged items are then reviewed by a human specialist, who decides whether any action is required. The first phase of the workflow is what Modl.ai calls “building game knowledge.” Developers upload assets—artwork, design documents, level layouts—and converse with the AI using natural‑language prompts. Holmgard notes that the interaction can include text, images, and other media, allowing the system to develop a contextual understanding of the game’s mechanics, narrative, and visual style. Once the AI has a baseline comprehension, developers specify the testing objectives.
These can range from classic, scripted test cases to more exploratory scenarios that are specifically crafted to exploit the strengths of the AI engine. Holmgard emphasizes that human expertise remains crucial at this stage: “You need to shape the prompts and language so the AI knows exactly what to do.” During a test run—referred to by Modl.ai as a “session”—the AI captures a wealth of data: textual logs, video snippets that serve as evidence, and comprehensive telemetry. This package is delivered to a human QA analyst, who reviews the findings, validates the bugs, and determines the appropriate next steps for the development pipeline. One of the persistent challenges in automated testing is distinguishing a genuine defect from an intended feature.
Holmgard explains that feedback loops from human testers help the AI refine its judgment. Over time, as testers label findings as “bug,” “not a bug,” or “expected behavior,” the system learns the nuanced design decisions that differentiate a flaw from a purposeful element.
Side envisions a future where QA professionals evolve into AI engineers. By upskilling staff to manage and improve the AI tools, Side hopes to move jobs up the value chain, offering higher‑pay roles and potentially repatriating work that had migrated to lower‑cost regions. McBride says, “Instead of a career that ends at senior tester, our people can become AI engineers, earn more, and open new career pathways.” This shift could also bring jobs back to markets where they were previously lost due to cost pressures.
From a day‑to‑day perspective, the AI system is designed to offload repetitive, low‑value tasks—such as smoke tests and routine verification—so that human testers can focus on strategic analysis, creative testing, and deeper gameplay evaluation. McBride notes that by automating the monotonous work, testers experience a more fulfilling career, spending their time on tasks that foster growth and innovation. Benedikte Mikkelsen, Modl.ai’s chief product officer, echoes this sentiment: “There are very tedious tests that don’t need a human to repeat over and over. Freeing testers from those allows them to assess gameplay quality or conduct more inventive testing, which is the real value proposition of AI.” Both executives stress that AI is not intended to replace QA staff but to augment them.
The price of manual testing in certain regions remains low, making a full AI takeover impractical. However, providing testers with AI‑enhanced toolsets extends their careers, improves accuracy, and reduces human error. McBride sums up the philosophy: “There will always be a human in the loop—that’s our ethos. AI helps us reduce the tedious workload and the mistakes that come with it, raising the overall quality of the process.” When asked about the types of games the solution currently supports, Chief Operating Officer Oliver Deslandes admits that the technology is best suited for simpler titles—puzzle games, card games, and other experiences with repeatable mechanics.
Fast‑paced 3D open‑world or fighting games are still a ways off. Deslandes explains that the initial focus is on clients willing to pilot the technology and accept its current limitations while providing valuable feedback for future enhancements. In summary, the Side‑Modl.ai partnership aims to blend human expertise with AI efficiency, creating a hybrid QA workflow that reduces repetitive labor, improves bug detection, and opens new career trajectories for testers.
The six‑month MOU will serve as a proving ground for this approach, with the hope that, as the AI matures, it can handle increasingly complex game environments while always keeping a human reviewer in the decision loop.