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 partnership aims to create a joint AI‑driven quality‑assurance (QA) platform that blends Side’s deep testing expertise with Modl.ai’s cutting‑edge machine‑learning capabilities.
Modl.ai, a relatively new player in the gaming sector, was founded in 2018 and secured $1.7 million in seed capital two years later. While the company has already experimented with AI‑enabled QA tools, earlier versions required developers to hand over entire game builds or even source code so that the AI could run its analyses.
This "integration‑based" approach, which relied on plugins and SDKs, proved cumbersome and created friction for developers, as Modl.ai’s founder Christoffer Holmgard acknowledges. The new solution sidesteps those hurdles by treating the game exactly as a human tester would receive it—delivered as a finished build that can be installed on a device identical to the end‑user’s hardware. The AI then watches the game play out on screen, simulates key actions, records every visual and data point, and passes the resulting dataset to a second AI layer that flags potential defects.
Those flagged items are then presented to a human QA specialist for final verification and prioritisation. Holmgard describes the first phase of the workflow as "building game knowledge." Developers upload assets, design documents, screenshots, and other relevant materials, then converse with the AI in plain language to teach it the fundamentals of the title. "All of this is in natural language," he says.
"We just chat with the system about our game, but we also feed it images, design docs, and other context." Once the AI has a solid grasp of the game world, developers specify the testing objectives. These can range from classic, scripted test cases to more exploratory scenarios that are tailored to the strengths of an AI‑driven engine. Holmgard stresses that human expertise is still required at this stage to craft effective prompts and guide the AI toward the right behaviours.
During a test run—referred to by Modl.ai as a "session"—the system records textual logs, short video clips that serve as evidence, and comprehensive telemetry. This package is handed over to a senior QA analyst, who reviews the findings, decides which issues merit further action, and adds contextual feedback such as "this is a bug," "this is an intended feature," or "this is not a bug because of X reason." Over time, the AI learns from this human‑provided feedback, refining its ability to distinguish between genuine defects and design choices.
One of the most challenging aspects for any automated tester is discerning whether a quirky behaviour is a flaw or a deliberate feature. The iterative loop of prompting, game‑knowledge accumulation, and human feedback helps the AI develop a nuanced understanding of game design conventions. Side sees this collaboration as a pathway to elevate its QA workforce.
Instead of remaining in purely repetitive testing roles, staff can transition into "AI engineers" who fine‑tune prompts, interpret AI outputs, and manage the overall testing pipeline. Martin McBride, Side’s chief commercial officer, explains that this shift moves employees up the value chain, offering higher salaries, new career trajectories, and the possibility of bringing jobs back to regions where low‑cost testing had previously been outsourced. McBride adds that AI will free testers from monotonous smoke‑test cycles that must be repeated for every new build. By automating those routine checks, QA professionals can focus on strategic analysis, deeper gameplay evaluation, and creative testing methods—ultimately leading to more fulfilling careers.
Benedikte Mikkelsen, Modl.ai’s chief product officer, echoes this sentiment, noting that many current tests are tedious and could be better performed by humans assessing gameplay quality rather than repeating identical flows. Both executives are clear that AI is not intended to replace human QA workers but to augment them. McBride points out that in some low‑cost regions, the price of manual testing remains so low that a complete AI takeover is impractical.
However, providing advanced tooling enables staff to extend their careers, take on more sophisticated responsibilities, and remain valuable to the organization. Side’s philosophy, according to McBride, is that AI operates on a "sliding scale"—the degree of automation can be adjusted, but a human will always be in the loop.
This ethos underpins the partnership’s goal of combining the precision and speed of AI with the judgment and creativity of experienced testers. When asked about the types of games the system 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 present greater challenges and are slated for future development.
Deslandes emphasizes that early adopters will likely be clients willing to pilot the technology and accept the learning curve associated with a nascent AI solution. In summary, the Side‑Modl.ai MOU represents a strategic attempt to fuse human expertise with AI efficiency in the realm of game QA.
By leveraging a non‑intrusive testing approach, building game knowledge through natural‑language interaction, and maintaining a human‑in‑the‑loop review process, the collaboration aims to reduce repetitive workload, improve defect detection, and create new, higher‑value roles for QA professionals. The journey is just beginning, but the partners believe that the combined strengths of their respective teams will gradually expand the scope of AI‑assisted testing across a broader array of game genres.