Can AI Review Games? — The Emergence of Reviews Without 'Experience'

In March 2026, a review of "Resident Evil Requiem" was briefly registered on Metacritic by UK gaming media Videogamer under the byline "Brian Merrygold" — then disappeared. When readers searched this name, nothing appeared: no online history, no past articles, no evidence of existence. Gfinity journalist Andrés Aquino first flagged this on X; Kotaku, Engadget, PC Gamer subsequently confirmed independently. The profile photo filename: "ChatGPT-Image-Oct-20-2025-11_57_34-AM-300×300.png"; the profile bio described him as a "seasoned analyst in iGaming and sports betting" — a gambling site specialist, not a game journalist. After exposure, the filename and bio were modified, but multiple outlets had already confirmed and reported the original. The deeper question this surfaces: AI can perfectly replicate the form of a review — it knows genre language, can imitate evaluation structure, can produce plausible sentences. So what makes a review a review? The answer is experience: time spent holding the controller, the physical memory of repeatedly failing at a specific section, the lingering impression that stays in your mind after the game ends. These cannot be generated by learning from published materials, however extensive. Experience can be summarized but cannot be replicated. But the truly important question is not "can AI write reviews without experience?" — it''s "have we ever actually verified that the reviews we read were backed by genuine experience?" The name of a known writer, a media logo attached — we assumed actual play was behind it. That assumption was never structurally verified. The Brian Merrygold incident reveals how fragile that trust foundation always was. The implication for game criticism: if audience trust in reviews depends on assumed experience that cannot be verified, the discovery that AI-written reviews are indistinguishable in form creates a verification problem that requires structural solutions — not just policy declarations from media organizations.