Unity: The (Perceived) AI Loser That's Quietly Inflecting

Unity hit $210 in November 2021. By August 2024, it traded at $13.89 — a 93% collapse that erased over $50 billion in market value. In January 2026, when Google unveiled Project Genie and the stock cratered another 24% in a single day, the consensus was clear: Unity was a perceived AI loser, a relic of the pre-generative era, a company that had destroyed itself through a toxic combination of developer betrayal, a disastrous acquisition, and an inability to find its place in the AI revolution.

That consensus is wrong. And the evidence is hiding in plain sight.

How Unity Broke Everything

The destruction was methodical and largely self-inflicted.

In July 2022, Unity announced a $4.4 billion all-stock merger with ironSource, an Israeli ad-tech company whose previous product, InstallCore, had been flagged by Windows Defender and Malwarebytes as malware. The developer community was already wary. Then CEO John Riccitiello called developers who resisted aggressive monetization "fucking idiots" in a trade press interview. The phrase became a rallying cry on Reddit's Unity subreddit.

The worst was yet to come. In September 2023, Unity introduced a runtime fee — charging developers up to $0.20 per game install. The policy was retroactive, meaning games already shipped would suddenly owe Unity money for downloads they'd already earned. Developers revolted. Studios publicly threatened to leave the engine. The open-source Godot engine saw its download numbers spike.

Riccitiello resigned in October 2023. Unity walked back the fee, but the damage was existential. Trust — the currency of a developer platform — had been spent.

Through 2024, the bleeding continued. Unity laid off 1,800 employees in January (25% of its workforce), then cut further in subsequent rounds. Revenue declined. Net losses hit $826 million in 2023. Nearly all key ironSource executives departed. A $4.4 billion acquisition was quietly disintegrating from the inside.

The AI Loser Label

If the self-inflicted wounds weren't enough, the market delivered another blow. On January 30, 2026, Google publicly launched Project Genie — a world model capable of generating interactive 3D environments from text prompts. Unity crashed 24% in a single session. Roblox fell 13%. Take-Two dropped 8%.

The thesis was simple and frightening: if AI can generate entire game worlds from a prompt, who needs a game engine? Unity was suddenly positioned on the wrong side of the most powerful technology trend in a generation. The stock hit a 52-week low of $15.33 in late February 2026, down over 92% from its peak.

The "AI loser" label stuck because it fit a convenient narrative. Unity was the company that couldn't get anything right — not its pricing, not its acquisitions, not its relationship with developers, and now, apparently, not its positioning against the AI future.

What the Market Missed

While the stock collapsed and the narratives compounded, something was changing inside the company.

Matt Bromberg became CEO in May 2024 — a turnaround specialist who had previously helped rescue Zynga. His playbook was unglamorous: stop the bleeding, cancel the runtime fee entirely (September 2024), rebuild developer trust, and bet the advertising business on an AI-powered platform called Unity Vector.

Vector is not a chatbot. It's not a world generator. It's a machine learning system that optimizes how mobile game ads are targeted and served — the kind of AI that doesn't generate headlines but does generate revenue. And the results have been extraordinary.

Here's what the filings show:

Revenue reversed from decline to double-digit growth in five quarters. Total revenue went from $435 million in Q1 2025 (down 6% year-over-year) to a preliminary $506 million in Q1 2026 (up 17%). Every single quarter accelerated: -6%, -2%, +5%, +10%, +17%.

Vector is the engine. Unity's AI-powered ad platform grew from an estimated $141 million quarterly revenue in Q2 2025 to approximately $217 million in Q1 2026 — a 54% increase in three quarters. It now represents roughly 62% of Unity's advertising revenue and is running at approximately $870 million annualized, tracking toward management's $1 billion target by year-end.

Margins are expanding, not just revenue. Adjusted EBITDA margins went from 19% to 26% over five quarters — a 700 basis point expansion. Q1 2026 EBITDA of $130-135 million beat guidance of $105-110 million by $20-25 million, the largest beat in the company's recent history.

Management has beaten guidance every quarter. Five consecutive revenue beats, ranging from $13 million to $21 million above the top of the guided range. Either management is systematically sandbagging, or they're genuinely surprised by Vector's performance. Either way, the pattern is consistent.

The ironSource Burial

Perhaps the most telling development: Unity is now unwinding the ironSource acquisition that nearly destroyed it.

On March 26, 2026 — alongside the Q1 preliminary results — Unity announced it would sunset the ironSource Ads Network entirely by April 30 and has engaged a financial advisor to divest Supersonic, ironSource's game publishing business. The $4.4 billion acquisition is being systematically dismantled less than four years after it closed.

But here's the number that matters: Unity's "strategic" revenue — excluding ironSource and Supersonic — grew 34% year-over-year in Q1 to $431 million. The advertising segment alone, stripped of ironSource, grew 48%. These are the growth rates of the real company, unburdened by the legacy of a deal that should never have happened.

The roughly $76 million in quarterly non-strategic revenue being removed was low-margin and declining. Its removal makes the company grow faster, earn higher margins, and simplifies the narrative from "turnaround story weighed down by a bad acquisition" to "AI-powered growth company."

The Irony of the AI Narrative

The deepest irony of Unity's "AI loser" label is that Vector — a real AI product shipping real results — has been driving the company's turnaround the entire time. While the market panicked about theoretical AI threats to game engines, Unity was deploying AI where it actually generates revenue: in advertising optimization.

Vector delivers a measurable 15-20% lift in installs and in-app purchase value for advertisers. Better AI means better returns for advertisers, which means they spend more, which generates more data, which makes the AI better. It's the kind of self-reinforcing loop that compounds quietly while the market debates hypothetical disruption.

When Bromberg responded to the Genie panic by saying world models are "complementary, not duplicative" to game engines, he was right — but he also could have pointed to his own AI story. Unity isn't losing to AI. It's winning with AI. Just not the kind of AI that generates headlines.

What to Watch

The inflection is real, but the next chapter is harder. With ironSource and Supersonic gone, there are no more legacy excuses. Unity needs to sustain 34% strategic revenue growth as easy comparisons from the 2024 trough fade. Vector needs to keep compounding as it scales past $1 billion — the same law of large numbers that's decelerating AppLovin's AXON (from 77% to ~52% year-over-year growth) will eventually apply.

The Create engine business — still the foundation of the ecosystem — grew 14% year-over-year on a strategic basis, healthy but not explosive. Unity remains deeply unprofitable on a GAAP basis, and still carries $2.2 billion in convertible debt.

But for a company that was left for dead — stock down 93%, developer trust shattered, flagship acquisition being dismantled, and labeled an existential AI casualty — the evidence in the filings tells a different story. Five quarters of accelerating revenue. Seven quarters of margin expansion. An AI product growing 15% sequentially in a seasonally weak quarter.

The market priced Unity for obsolescence. The filings show inflection.