Side, a seasoned services provider formerly known as PTW, 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 venture that will blend Side’s extensive QA expertise with Modl.ai’s cutting‑edge AI technology to create a next‑generation, AI‑enhanced quality‑assurance platform for the gaming industry. Modl.ai, a relatively new player in the games sector, was founded in 2018 and secured $1.7 million in seed financing two years later.
According to Martin McBride, Side’s chief commercial officer, the partnership “marriages the best of Side’s QA know‑how with the most advanced AI capabilities Modl.ai has built.” He adds that the collaboration aims to deliver what many in the industry have called the "holy grail" of QA: a seamless, AI‑powered system that augments human testers rather than replaces them. Historically, AI‑driven QA tools required developers to hand over complete game builds or even source code, often via plugins or SDKs, so the AI could analyze internal structures. This "integration‑based" approach proved cumbersome and created friction for developers, as noted by Modl.ai founder Christoffer Holmgard. The new solution abandons that model.
Instead, the AI receives the same binary a tester would use, installs it on a device that mirrors the consumer’s hardware, and then runs the game while observing the screen, simulating inputs, and logging every event. A second AI layer reviews the generated data, flags potential defects, and escalates them to human reviewers for final judgment. The first phase of the workflow is what Modl.ai calls "building game knowledge." Developers upload assets, design documents, screenshots, and other reference material, then converse with the AI in plain English to teach it the game’s mechanics, narrative, and visual language.
Holmgard explains that this natural‑language interaction helps the system form a contextual model of the title, which later informs its testing strategies. Once the AI has a solid understanding, developers define test objectives.
These can range from traditional test cases—such as verifying menu navigation or level loading—to novel scenarios crafted specifically for AI execution. Holmgard stresses that human expertise is still essential at this stage to phrase prompts correctly and guide the AI toward meaningful actions. During a test run, the system captures textual logs, short video clips (referred to as "evidence"), and comprehensive telemetry. These artefacts are passed to a human QA specialist, who reviews the findings, decides which issues merit further action, and provides feedback to the AI.
This feedback loop helps the AI differentiate between genuine bugs and intentional design features, a nuanced distinction that purely algorithmic approaches often miss. Over time, as testers annotate bug reports—labeling items as "bug," "not a bug," or "feature with justification"—the AI refines its internal models and becomes more attuned to the particular design philosophy of each studio.
Holmgard notes that this iterative learning process, combined with the initial natural‑language knowledge base, enables the system to improve its accuracy and reduce false positives. Side sees broader strategic implications in this technology. By equipping its QA staff with AI tools, the company can elevate their roles from repetitive test execution to "AI engineers" who design prompts, interpret AI‑generated insights, and shape testing strategies.
McBride argues that this shift moves employees up the value chain, offering higher salaries, new career pathways, and the possibility of reshoring jobs that were previously off‑shored due to low testing wages. According to McBride, the AI also frees testers from monotonous smoke‑test cycles that accompany every new build. With the AI handling routine verifications, human testers can focus on strategic analysis, gameplay assessment, and creative testing approaches—tasks that foster professional growth and job satisfaction. Benedikte Mikkelsen, Modl.ai’s chief product officer, echoes this sentiment, emphasizing that eliminating tedious repeatable flows lets humans concentrate on higher‑order evaluation of game feel and design.
Both leaders are clear that AI will not replace QA personnel. Instead, it will act as a force multiplier, allowing testers to apply their expertise where it matters most. McBride points out that in regions where testing costs are extremely low, a complete AI takeover is neither feasible nor desirable; the human element remains indispensable for nuanced judgment and for maintaining a long‑term talent pipeline. Side’s philosophy frames AI as a "sliding scale"—a spectrum where automation handles the low‑value, high‑volume tasks while humans intervene on complex, judgment‑heavy issues.
"There will always be a human in the loop," McBride says, describing this as the company’s core ethos. By removing human error from repetitive steps, the partnership aims to raise overall quality, reduce mistake rates, and accelerate release cycles.
When asked about the types of games the solution currently supports, COO 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 remain a longer‑term goal. Deslandes stresses that early adopters need to be willing to pilot the system and accept that the technology will evolve alongside their feedback. In summary, the Side‑Modl.ai collaboration seeks to blend AI efficiency with human insight, creating a QA workflow that is faster, more accurate, and more rewarding for the people who run it.
The six‑month MOU will serve as a proving ground for this vision, with the hope that, as the AI matures, it will handle an ever‑broader range of testing scenarios while always keeping a human reviewer at the helm.