Side, a seasoned services provider, has entered into a six‑month memorandum of understanding with Denmark‑based artificial‑intelligence specialist Modl.ai to co‑develop an AI‑driven quality‑assurance platform for video games. The partnership merges Side’s deep‑rooted expertise in testing services—formerly operating under the PTW brand—with Modl.ai’s newer, data‑centric approach to game testing. Modl.ai, founded in 2018, secured $1.7 million in seed financing two years later, positioning itself as an innovative player in the gaming ecosystem. Martin McBride, Side’s chief commercial officer, describes the MOU as a union of the strongest assets each company brings to the table.
“One of the most coveted, almost holy‑grail, goals for the games industry today is to harness AI to amplify QA capabilities. Nobody has nailed a robust solution yet.
We believe we have a real shot at it,” he says. Earlier AI‑enabled QA tools—including Modl.ai’s initial offering—required developers to ship full builds or even share source code with the testing vendor.
That approach relied on integration plugins and SDKs, creating considerable friction and, as founder Christoffer Holmgard notes, becoming “hard to sustain over time.” The new system eliminates the need for internal access. Instead, developers provide the same product they would hand to a human tester or eventual consumer, loading it onto a comparable device. The AI then watches the gameplay, simulates key actions, records every visual and input event, and passes the resulting data set to a second AI layer that flags potential defects.
Human reviewers then decide whether each flagged issue warrants action. The first phase, dubbed “building game knowledge,” involves developers uploading assets—images, design documents, even conversational snippets—so the AI can develop a natural‑language understanding of the title.
“All of this happens in plain language,” Holmgard explains. “We chat with the system about the game, feed it visual references, and feed it design paperwork.” Once the AI has a contextual map of the game, developers specify the testing scope. This can range from classic test cases to novel scenarios tailored to the AI’s strengths. Holmgard stresses that human expertise remains essential at this stage to craft effective prompts and guide the AI’s behavior.
The system then executes a “run,” generating bug reports enriched with text descriptions, video clips (the so‑called “evidence”), and comprehensive log data. These artifacts are reviewed by seasoned QA specialists, who determine the appropriate next steps for the production pipeline.
Distinguishing between a genuine bug and an intentional feature is a nuanced challenge for any automated system. Holmgard notes that continuous feedback from human testers refines the AI’s judgment. “Prompt engineering, pre‑built game knowledge, and iterative tester feedback all help the model learn what constitutes a defect versus a design choice,” he says.
Over time, the AI becomes attuned to the specific quirks of each title. Side envisions a future where QA personnel transition into AI‑engineer roles, thereby reshaping the global distribution of testing jobs. McBride points out that this shift can elevate staff up the value chain, offering higher salaries, new career trajectories, and the possibility of bringing work back to regions that previously lost jobs due to cost pressures. “When our people become AI engineers, they can command better pay, and we can repatriate talent to markets where it once vanished,” he explains.
The AI tool also promises to make QA careers more rewarding by automating repetitive, mundane tasks. McBride observes that each new build typically requires smoke tests and numerous verification steps that can bog down testers. By offloading these routine checks to AI, testers can focus on strategic analysis, creative problem‑solving, and deeper gameplay evaluation.
Benedikte Mikkelsen, Modl.ai’s chief product officer, adds that eliminating tedious, repetitive flows frees humans to assess core gameplay and conduct more inventive testing, which she sees as the primary value proposition of AI‑augmented QA. Both executives stress that AI is not a replacement for human QA workers but a catalyst for professional growth. McBride notes that testing costs remain low in some regions, making a complete AI takeover impractical.
Instead, providing sophisticated tools enables staff to extend their careers, move into higher‑skill positions, and reduce error rates associated with monotonous manual testing. “Our ethos is that there will always be a human in the loop,” he says, describing AI as a “sliding scale” that complements, rather than supplants, human judgment.
Regarding the types of games the solution currently supports, COO Oliver Deslandes admits 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 remain a longer‑term goal. Deslandes emphasizes that early adopters should be willing to embark on a collaborative journey, recognizing that the platform will evolve based on their feedback and requirements. “It’s a journey,” he concludes.
“If you’re willing to pilot with us now, you’ll see how the technology matures and where it can take us together."