Masahiro Sakurai, the designer behind Kirby and Super Smash Bros., says generative AI will push the number of games released on Steam well past the 19,000 titles he cited, and he expects most of the extra releases to become “unprofitable titles.” The Tribune’s report on the interview frames it as a warning about originality, and headline writers elsewhere have boiled it down to a simpler claim: AI will flood the games industry with generic titles.
Sakurai made the comments in an interview with the Japanese outlet Denfaminicogamer. The story has since spread through LevelUp, shattered.io and OtakuKart, among others. It lands while AI inside game studios is already a flashpoint, as our report on an Xbox AI strategist deleting their account within 24 hours showed. Sakurai’s version of the argument is different from the usual ethics fight, though. He is talking about volume, sameness and money.
This analysis separates what Sakurai said from what headlines added, explains why a Masahiro Sakurai AI warning carries weight with developers, and looks at who actually pays when a storefront fills up with look-alike games. It also maps the likely market effects and offers four forecasts. Where the record is thin, the article says so rather than filling the gap.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
What Masahiro Sakurai Said About AI and Steam
The core of the interview is short. Sakurai pointed to Steam, which had reached 19,000 titles in the year he discussed, and said that figure will keep rising because AI will be used to produce games. He then predicted that most of those games will end up unprofitable, and that there will be so many that players will struggle to track or understand them.
He also made a point about design. According to the coverage, he argued that AI-generated games, and games mass-produced with little care, will mostly reproduce systems players have already seen in other games. That is a claim about creative range, not about technical quality. A game can run perfectly and still add nothing.
The 19,000-title figure
The Steam number is the only hard statistic in the story, and it comes from Sakurai himself as relayed by the press. No outlet in the coverage we reviewed supplied an independent monthly breakdown, so treat 19,000 as his stated benchmark for Steam’s annual output rather than an audited count. Its value is rhetorical. It tells readers the shelf is already crowded before any AI-driven growth arrives.
Why he expects the number to keep climbing
The logic is about cost. Generative tools lower the effort needed to turn out a playable build, so more people can ship more games. Sakurai did not name a specific AI tool, company or product in the reported remarks, and neither will this article. The argument stands on the incentive alone: when making games gets cheaper, more games get made.
What Is Confirmed and What Is Not
Several details in circulating headlines go further than the record supports. The table below sorts the claims by how well the coverage backs them.
| Claim | Status | What the coverage shows |
|---|---|---|
| Sakurai spoke to Denfaminicogamer about generative AI and game development | Confirmed | Reported across multiple outlets |
| Steam reached 19,000 titles in the year discussed | Confirmed (as Sakurai’s figure) | Relayed by Tribune and others |
| He expects the number to keep rising because of AI | Confirmed | Consistent across reports |
| Most such games will be “unprofitable titles” | Confirmed | The most consistently reported wording |
| Too many games for people to track or understand | Confirmed | Consistent across reports |
| AI or low-care mass production mostly reproduces known systems | Confirmed | Reported by VGC and others |
| Most AI games will be “unplayable” | Unconfirmed | Appears in some reports only |
| AI will “flood the games industry with generic titles” | Unconfirmed as a quote | A headline summary, not his words |
| A specific AI tool, company, budget, price or date | Unconfirmed | None named in the reported remarks |
The gap between “unprofitable” and “unplayable” matters. One is a market forecast, the other a quality verdict, and they lead to different conclusions. The wording reported most consistently is unprofitable, so that is the one this article treats as his.
Why Generic Games Are an Economics Problem
Read closely, Sakurai’s warning is about attention, not aesthetics. A player has a fixed number of evenings in a year. If the pool of releases doubles and the evenings stay put, the average game earns less. That holds even when every new game is competent.
He describes the likely result in a single phrase: unprofitable titles. A studio that spends money on a game, then watches it vanish into a catalog nobody can browse, fails regardless of quality. The same flood dynamic has hit other fields. Our report on AI speeding up coding while flooding open source with slop shows the pattern in software, where cheap generation raised output faster than anyone’s ability to review it.
Games add a wrinkle. Software maintainers can reject a bad pull request. Players cannot preview a hundred hours of design before buying. They rely on trailers, reviews and recommendation systems, all of which strain when supply jumps. That strain is the discoverability problem Sakurai describes when he says people will not be able to keep track of or understand what is out there.
AI-Generated Games and the Originality Argument
Sakurai’s second claim is the more interesting one. AI-generated games, he said, tend to repeat systems already seen in other games. He told the interviewer that examples of this already exist today. This follows from how generative tools work. They learn from what has shipped, so their default output sits close to what has shipped.
He put the distinction plainly, according to GoNintendo’s coverage: “Basically, AI refers to things that have been created so far, but it can’t refer to things that haven’t been created before.” The sentence is not a prediction about technology. It is a claim about where new ideas come from, and it places that source with people.
Critics of this view will point out that humans also copy. Most genres are built on shared mechanics, and every studio borrows. Sakurai does not dispute that. His target is mass production “with little care,” where copying is the whole method rather than a starting point. The difference is intent, and intent is hard for a storefront to measure.
Ico, Shadow of the Colossus and the Fumito Ueda Reference
The cited coverage brings Fumito Ueda into the discussion, along with Ico and Shadow of the Colossus. Those games work as a shorthand for authored vision: small in scale, distinctive in mood, and hard to describe as a mix of earlier designs. Sakurai uses them to illustrate what originality looks like and where he sees AI’s limits.
Be careful not to stretch this. The coverage does not say Ueda commented on Sakurai’s remarks, and it does not say Ueda endorses or rejects any AI tool in this story. The reference is Sakurai’s, offered as an example. The point for readers is that a famous designer reached for two decades-old games to explain what a statistical model cannot easily supply: a fresh artistic direction.
Sakurai in His Own Words
Three direct quotes from the published coverage show the shape of his argument. On profitability, Eurogamer’s report carries his forecast for the added releases: “I think most of them will become ‘unprofitable titles.’” He is describing an outcome for the group, not a verdict on any single game.
On sameness, Video Games Chronicle quotes him saying: “When it comes to AI-generated games, or those churned out through mass production with little care, I think most of them will simply replicate systems that we’ve already seen in other games. There are already plenty of examples of this now.” The second sentence is the one worth underlining. He treats the problem as present, not hypothetical.
On the human side, the GoNintendo quote above frames creativity as the part of the job that points at things that do not yet exist. Together the three lines make a coherent position. More supply, less profit per title, and a premium on people who can imagine something new. Nothing in the quotes calls for banning a tool.
Market Impact: Who Pays When the Storefront Floods
A bigger catalog does not hurt everyone equally. The table below is editorial analysis of how Sakurai’s scenario would spread cost across the industry. It contains no new statistics.
| Group | Likely effect of a flooded storefront | Main risk |
|---|---|---|
| Players | More choice, harder search | Wasted money and time on look-alikes |
| Solo and small indie developers | Lower barrier, steeper visibility climb | Good games buried under volume |
| Mid-size studios | Pressure to stand out with distinct design | Budgets that assume steady sales |
| Large publishers | Brand and marketing reach become moats | Pressure to cut cost with the same tools |
| Platform holders | Higher revenue from sheer listings, weaker trust | Customer fatigue and reputational damage |
| Critics and streamers | More demand for filtering and curation | Overload and burnout |
Smaller markets feel visibility problems first. Our look at how Indonesia’s local game developers hold only a sliver of their home market shows how hard it already is for regional studios to be seen when global releases dominate. Add a wave of cheap, near-identical games and that climb gets steeper.
Large publishers face their own version. Cost pressure at the top of the industry is real, and our analysis of the BioWare closure and EA’s economics shows how quickly a long-running studio can lose its place when the math turns. If AI cuts the cost of mid-tier content, some of that pressure will push publishers to use it, which feeds the very supply Sakurai worries about.
Historical Context: Gluts Have Hit Games Before
The industry has lived through oversupply before. The North American video game crash of 1983 is often tied to a glut of weak releases that left retailers and buyers wary of the whole category. The details are debated, but the shape is familiar: too many similar products, falling trust, and a market that took years to recover.
Digital storefronts have since cycled through their own versions. Open submission models invite bursts of low-effort uploads, and platform holders respond by adjusting gatekeeping. Valve itself has changed how Steam approves games over the years, moving between looser and tighter systems. The pattern suggests a platform can survive a glut, but only after it builds better filters.
What is new is the speed. Earlier gluts needed teams, publishers or at least a lot of free evenings. Generative tools shrink that effort, so the same dynamic can build faster than platform policy can respond. That is the real novelty in Sakurai’s warning. Our piece on GTA VI’s budget and the limits of AAA gaming shows the opposite pole, where a handful of enormous productions soak up attention while everything else fights for the remainder.
Storefront Models Compared
How exposed a platform is depends on how it lets games in. The table compares three broad models. It is a general framework, not a claim about any one company’s current policy.
| Model | How games get listed | Exposure to AI-driven volume | Main strength |
|---|---|---|---|
| Open submission PC storefront | Developers publish with light review | High | Huge variety and low barriers |
| Curated console store | Platform approval before listing | Medium | Quality floor and trust |
| Subscription catalog | Platform selects titles for the service | Low | Editorial control and bundled value |
| Editor-led or community lists | Third parties recommend games | Low | Filters that match taste |
Sakurai chose Steam as his example, and the reason is plain. An open storefront takes in the most volume, so it feels the pressure first. Platforms with approval steps or editorial selection gain a selling point if floods arrive: they can promise that someone has already looked.
What Developers and Publishers Can Do
The practical reading of Sakurai’s remarks is a call to compete where AI is weakest. If generated games copy known systems, then a new system, a sharp visual identity or an unusual structure becomes the differentiator. Studios that can describe what is new about their game in one sentence will have an easier time reaching players than those who say it is like three other games but better.
Teams should also invest in discovery. Wishlists, demos, community channels and creator relationships do the filtering work that storefront browsing used to do. None of this is glamorous, and all of it grows more valuable as the shelf fills up.
Using AI tools is not the problem Sakurai describes. His concern is mass production with little care. A studio that uses a tool for a narrow task and still puts an authored idea at the center is outside the scenario he warns about. The line he draws runs between thoughtful use and volume for its own sake.
Four Predictions for AI and the Games Market
These are editorial forecasts, not reported facts. They follow from the incentives Sakurai describes and carry real uncertainty.
- Discovery tools become a selling point. Platforms that help players filter out look-alikes, through better labeling, ranking or human curation, will market that as a feature.
- Trusted names gain pricing power. Designers and studios with a recognizable voice will rely on reputation as the shortcut that a crowded catalog cannot provide.
- The debate shifts from tools to taste. Arguments over whether AI was used will give way to a harder question about whether a game offers something a player has not already done.
- More named creators will weigh in. Sakurai’s interview adds weight to a conversation that other well-known designers are likely to join, which will keep the topic in the news cycle.
What to Watch Next
Three signals will show whether the warning is playing out. First, watch whether storefront release counts keep growing and whether independent analysts publish clearer monthly data than the single figure now circulating. Second, watch for platform policy changes aimed at low-effort uploads. Third, watch sales concentration. If a shrinking share of releases takes a growing share of revenue, Sakurai’s unprofitable-titles forecast is arriving.
We will keep tracking the story as more coverage appears. For wider context on games, platforms and the business behind them, see the gaming section of Tech Insider.
Frequently Asked Questions
Who is Masahiro Sakurai?
Masahiro Sakurai is a Japanese video game designer best known as the creator of Super Smash Bros. and Kirby. Coverage of this story describes him in those terms.
What did Sakurai say about AI and Steam?
In an interview with Denfaminicogamer, he said Steam had reached 19,000 titles in the year discussed and that the number will keep growing because AI will be used to produce games. He predicted most of those titles will become unprofitable and said there will be too many for people to track or understand.
Did Sakurai say AI-generated games would be unplayable?
That wording appears in some reports, but it is not the most consistently reported version. Most coverage, including Eurogamer, uses “unprofitable titles,” so this article treats that as his phrase and the unplayable wording as unconfirmed.
Did he literally say AI will flood the industry with generic games?
No confirmed quote says that. The phrasing is a headline summary of his comments about rising volume and games that repeat existing systems.
Why are Ico and Shadow of the Colossus mentioned?
The coverage uses Fumito Ueda and those two games when discussing originality and the limits of AI. They stand in for authored, distinctive design that is hard to derive from earlier games.
Which AI tool or company was he talking about?
None. The reported remarks do not name a tool, company, product, budget or release date.
Is Sakurai calling for AI to be banned in games?
The published remarks do not say that. They focus on the number of releases, the sameness of AI-generated and low-care games, and the commercial outcome for most of them.
