Steam Vs. Epic | AI and the Great Debate

Tim Sweeney keeps framing Steam’s AI disclosure as a political scarlet letter. The policy is narrower than most people think, and the argument against it leaves the customer out entirely.

First, what the disclosure actually says, because almost nobody reads it. Steam’s third content survey covers generative AI, and the text is explicit. We are aware that many modern game development environments have AI powered tools built into them. Efficiency gains through the use of these tools is not the focus of this section. What it is concerned with is AI used to create content that ships with your game and is consumed by players. Artwork, sound, narrative, localization.

It splits into two buckets. Pre-generated content, which Valve evaluates the same way it evaluates any non-AI content. And live-generated content, created while the game is running, which carries one extra requirement: tell them what guardrails you have so it is not producing illegal content.

That is it. It is not every AI powered tool. It is not your coding assistant. Steam updated the language earlier this year specifically to make that clearer.

Sweeney’s position is that these labels make no sense because nearly all future production will involve AI, and that the real purpose is to drive cancel campaigns. He has gone further and framed it as a political act. I disagree with that wholeheartedly.

How people feel about AI is not a political issue. Go look at any survey. You will find people across the political spectrum on every side of this. It does not sort the way he is describing it.

And on the cancel mob argument, that is a non argument to me, because it happens either way. I have dealt with it in comics with nothing AI related anywhere near it. Put anything out and somebody will rally people against you. That is the cost of shipping work, not the cost of a label.

Here is what is missing from his read. He is looking at this purely as a developer and a competitor. He is not looking at it as a customer.

The data right now says roughly half of people view this technology negatively, for a lot of different reasons. Ethical ones. Economic ones. Some of them just do not want to consume it. You can think those reasons are illegitimate. They still exist, and the market has to be dealt with as it is, not as you wish it was.

Steam got where it is by being friendly to developers and friendly to customers. A disclosure that lets a buyer make an informed decision, with a text box where the developer explains how they actually used it, is a pro customer move. Plenty of buyers will read that and decide they are fine with it.

Hiding it is not an argument.

And a broader note for anybody bullish on this. You want to be taken as seriously as people who do not use it, but you are not ready to be criticized like them. Grok just put out an AI Odyssey and it is not good. If Hollywood released that, it would get laughed out of the room, and the answer cannot be that you have to judge it on where the technology is going. You judge it on where it is right now.

That criticism was coming either way. Everybody in business has dealt with it forever. AI is not immune to it and you should be prepared for it.

The verdict on all of this is still out. I am far more interested in the economics than the capabilities, because the funding is not based on what this does today. It is based on what it might do later. Do not jump the gun.