Side, a veteran 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 effort to build a next‑generation, AI‑enhanced quality‑assurance (QA) platform tailored for the games industry.
While Side brings decades of experience in outsourced services and a deep bench of QA talent, Modl.ai contributes cutting‑edge machine‑learning algorithms and a vision for how AI can augment, rather than replace, human testers. The partnership is framed as a "sliding scale" approach to automation, a phrase coined by Side’s chief commercial officer Martin McBride. "There will always be a human in the loop; that's our ethos," he explains. In practice, this means AI will handle the most repetitive, data‑intensive tasks, freeing human specialists to focus on strategic analysis, creative testing, and the nuanced judgment calls that only seasoned QA professionals can make.
Historically, AI‑driven QA tools required developers to hand over full builds or even source code so the system could instrument the game and run automated checks. Modl.ai’s earlier offering relied on an integration model that involved a plug‑in and SDK, creating friction for studios that were reluctant to expose internal assets.
Founder Christoffer Holmgard acknowledges that this approach "became difficult over time" because of security concerns and the overhead of maintaining tight integration pipelines. The new solution abandons that model entirely.
Instead, developers submit the same package they would give to a human tester—a compiled build ready for consumer use. The AI runs the game on a device that mirrors the target platform, watches the screen, simulates inputs, and records every observable event. This raw data is then fed into a second AI layer that scans for anomalies, flags potential bugs, and escalates findings to a human reviewer.
A crucial early step is what Modl.ai calls "building game knowledge." Teams upload assets, design documents, screenshots, and other reference material, then converse with the AI in natural language to teach it the game's mechanics, narrative flow, and expected behaviours. Holmgard notes that this dialogue can include text, images, or even video clips, allowing the system to develop a rich, multimodal understanding of the title before testing begins. Once the AI has a baseline comprehension, developers define test objectives. These can range from traditional test cases—such as verifying that a menu opens correctly—to more exploratory scenarios that leverage the AI’s ability to generate novel inputs.
"Human expertise is still needed to craft the right prompts and guide the AI toward useful actions," Holmgard says. QA engineers must translate their domain knowledge into clear, structured instructions that the AI can execute. During a test run, the system captures a variety of artefacts: textual logs, short video snippets that serve as visual evidence, and comprehensive telemetry data. All of this is compiled into a bug report that is handed off to a senior QA specialist.
The specialist reviews the findings, decides whether each issue is a genuine defect, a design choice, or a false positive, and provides feedback that the AI uses to refine its future predictions. Over time, this feedback loop helps the model distinguish subtle differences between intentional game features and unintended bugs. Side sees this iterative process as a pathway to upskill its workforce.
Rather than remaining in low‑margin testing roles, employees can evolve into "AI engineers" who manage prompts, interpret model outputs, and fine‑tune the system for specific titles. McBride argues that this shift not only raises salaries and career prospects but also creates the possibility of bringing jobs back to regions that have lost QA work due to cost pressures. By moving staff up the value chain, Side hopes to retain talent longer and diversify its geographic footprint. The human‑centric philosophy extends to the day‑to‑day experience of testers.
Repetitive smoke tests and regression checks, which can dominate a tester’s schedule, are now automated. This frees QA professionals to engage in higher‑order activities such as gameplay analysis, risk assessment, and creative test design. "When we knock out the boring, repetitive work, our people can think more strategically and feel more fulfilled," McBride explains.
Modl.ai’s chief product officer Benedikte Mikkelsen echoes this sentiment, noting that many current test procedures are monotonous and could be performed more efficiently by a machine. "The value proposition of AI is that it handles the tedious flows, allowing humans to focus on assessing gameplay, exploring edge cases, and delivering insights that only a seasoned player can provide," she says. Both companies stress that AI will never fully replace QA staff. The price point for manual testing in certain low‑cost regions remains so low that a complete automation takeover is neither feasible nor desirable.
Instead, the technology acts as a force multiplier, extending the capabilities of human testers and enabling them to stay relevant in an evolving industry. Regarding the types of games the solution can currently support, Side’s chief operating officer Oliver Deslandes is candid: the system excels with simpler, repeatable titles such as puzzle games, card games, and other experiences that do not demand ultra‑fast, physics‑heavy interactions.
"Fast‑paced fighting games or large open‑world adventures are still a ways off," he admits. The focus, for now, is on titles with predictable mechanics where the AI can reliably model player behaviour and detect deviations. Deslandes emphasizes that the partnership is a pilot journey. Early adopters willing to experiment will help shape the roadmap, providing real‑world data that informs future enhancements.
"If you’re at the table with us now, you’ll see where this technology goes as we iterate together," he concludes. In summary, the Side‑Modl.ai collaboration aims to blend the analytical power of AI with the contextual insight of human QA experts.
By automating routine testing, building a shared language around game knowledge, and creating new career pathways, the alliance hopes to redefine quality assurance for the modern gaming landscape while keeping the human element firmly at the core of the process.