Ex-Epic Director Bets AI Can Dethrone Unreal Engine

Eight and a half years inside Epic Games gives you a particular kind of perspective on Unreal Engine. You learn where the seams are, which architectural decisions from 1998 still dictate how things work in 2026, and exactly how painful it is to bolt new technology onto a codebase that was originally designed around mouse-driven menus. Arjan Brussee has that perspective, and he's now using it to build a competitor.

Brussee, who co-founded Guerrilla Games and served as Epic's global director of product management for Unreal Engine, told the Dutch podcast De Technoloog that he's developing a new engine called the Immense Engine. The pitch: a fully European-hosted platform built from the ground up around AI agents, designed to challenge American and Chinese engines that currently dominate the market.

"No one is currently making an engine that is fully European-hosted, built by Europeans, and complies with European rules and guidelines," Brussee said. He framed existing engines like Unreal as products of a different era, "made for and by people who have to click through a menu with a mouse. If you want to change something, it has to be done for the entire engine."

His proposed solution is modular AI. Rather than a monolithic stack where every update requires reworking the whole framework, the Immense Engine would use AI agents as building blocks, letting developers swap in new systems as the technology evolves. "If you are smart and know how to put a good framework of AI agents to work, you can do the work of ten or fifteen people," he added.

The Résumé Behind the Bet

Brussee's career reads like a tour of the industry's technical backbone. He programmed Epic's Jazz Jackrabbit games in the 1990s, co-founded Guerrilla Games in 2003, spent time at EA working on Battlefield titles, co-founded Cliff Bleszinski's Boss Key Productions, and then returned to Epic for over eight years across roles spanning mobile, product management, and technical direction. The man has seen how engines get built, maintained, and eventually outgrown. That experience is the entire basis of his argument: he knows what Unreal can't easily become, because he was there while it tried.

I find the framing interesting, even if the execution is going to be brutally difficult. Brussee isn't just pitching a lighter Unreal clone. He's arguing that the fundamental architecture of today's major engines is wrong for an AI-driven future, and that starting fresh is the only way to get it right. Coming from someone who spent nearly a decade inside Epic, that's not a throwaway claim.

But there's a canyon between diagnosing the problem and shipping the solution. Modern game engines represent decades of accumulated development work. Unreal Engine has an ecosystem of plugins, marketplace assets, documentation, and community knowledge that no startup can replicate in a few years. Unity, for all the self-inflicted damage from its 2023 install-fee fiasco, still powers a massive chunk of the indie and mobile market. And then there's Godot, the open-source engine that's been quietly absorbing developers who lost trust in Unity. Slay the Spire 2 runs on it. It's free, it's open, and it sidesteps the geopolitical concerns Brussee is raising without requiring anyone to buy into an AI-agent architecture that doesn't exist yet.

Brussee is also positioning the Immense Engine for non-gaming markets. He cited 3D simulations for defense and logistics as potential use cases, noting that European data compliance requirements make a locally hosted engine attractive to sectors where sending data through American or Chinese cloud infrastructure is a dealbreaker. That's a smart angle for funding and early adoption, even if it doesn't solve the core challenge of convincing game studios to abandon tools they already know.

A Real Gap or a Sales Pitch?

The European sovereignty argument has real weight in 2026. Trade tensions, data regulations, and a growing reluctance to depend on American tech platforms have created demand for homegrown alternatives across multiple industries. Whether that demand extends to game engines specifically is less clear. Most European studios I can think of are deeply embedded in Unreal or Unity workflows, and switching engines mid-pipeline is one of the most expensive decisions a studio can make. The Immense Engine would need to offer something so dramatically better that the switching cost feels justified, and "AI agents" as a concept is still far too vague to evaluate on those terms.

I've covered enough AI pitches in gaming over the past year to know that "AI-first" is becoming the new "blockchain-enabled" in terms of how freely it gets thrown around. Brussee has more credibility than most people making these claims, given his actual engineering background and his years inside the engine he's trying to replace. But credibility and a working product are different things, and there's no release timeline for the Immense Engine yet. Building a game engine from scratch is a multi-year, multi-hundred-million-dollar undertaking even with conventional architecture. Doing it with an unproven AI-agent framework adds a layer of technical risk that no amount of industry experience can fully mitigate.

According to Game Rant's reporting, the Immense Engine could potentially fill a gap left when Unity Technologies relocated from Copenhagen to San Francisco in 2009 to access American venture funding. Whether Brussee is pursuing EU funding for the project hasn't been confirmed.