Denuvo Know They're In Trouble.

Vilard will not have Denuvo on PC. We trust you. I don't know about you, but that statement was not in my 2024 EA Games Bingo, and it's something said by the executive producer of Veil Guard, Mike Gamble, just before the launch.

What it represents is a bit of a different attitude towards risk than making a different business decision, and I would say not actually them trusting you. Now, of course, everybody—EA included—knows that Denuvo has a bad name. Denuvo also knows they have a bad name, and for Denuvo, this is all quite terrible.

Their bad rep could actually be getting in the way of their business, so they know they're behind. What they've tried to do now is to respond to the situation. So, the drop to speed—Denuvo is an anti-piracy solution for PC games that's not just a simple "phone home" to a server to authenticate that the game is legitimate.

Sort of thing. What it does is it works at runtime using virtualization and other obfuscation techniques to prevent tampering with game code. Right now, there are times where this has had a barely noticeable impact on game performance, but not always.

There are well-publicized instances of Denuvo-protected versions of games running pretty damn badly and clearly documented cases where a cracked version of a game that removes Denuvo will significantly outperform the legally purchased Denuvo counterpart. Now, Denuvo had always held the line that that's not what's going on and that it's not their fault. But at the very least, what their pitch would be, right, is that, look, your game is going to be cracked at some stage, but the point of Denuvo is they will make it take so long for your game to be cracked that your launch window at full RRP will be protected.

As for Denuvo examples recently, well, we've got Black Myth: Wukong that had Denuvo, and Kingdom Come: Deliverance, too, which confirmed that it would have Denuvo on launch on PC, meaning that Steam forum users were not best pleased. So, that's the thing: Denuvo gets hate; their partners get hate. Now, EA have left them out, and they know they have a bad reputation.

To solve all this, they've decided that it's time to start talking directly to us and to the press and to support our work in stories like this. I've partnered with today's sponsor, where I found a grimly funny trio of AI stories: hospital AI failing, gun checking AI failing, and, uh, yeah, Nvidia blowing past Apple as the world's most valuable company. You see, I use Ground News—it's where I go to stay in the loop on the things I care about.

And you can get 40% off at ground. news/bellular. All those three stories really stuck out to me: the researchers investigating OpenAI's Whisper being implemented in medical centers just found wild inaccuracies and hallucinations.

Like, one researcher said he found hallucinations in eight out of the 10 transcripts that he checked, which is kind of wild. With Ground, I can actually see consensus, which is good for trust. I can see 45 sources; 10 are leaning to the left, six are leaning to the right, and others are in the center, and those are all ratings that come from three independent news monitors.

I can see their corporate ownership; I can see factuality ratings, which is so helpful. Now, where Ground gets powerful—and honestly, a little bit fun—is headline comparison. In this story, the sides are great, very much unlike this story of The Washington Post not endorsing a presidential candidate.

CNN ran with, "Washington Post won't endorse candidate in 2024 presidential election after Bezos decision. " OutKick went with, "Sorry, Kamala: Washington Post chooses not to endorse presidential candidate for first time in 36 years. " So, um, you can probably tell which one's leaning left and which one's leaning right there.

That's the power of Ground: you can see all of the stories' top levels and see all of the spin consolidated in one place, and that means that I've got the tools to see the whole picture so I'm not misinformed. I'm actually—it's in the name—well-grounded. You can join me at ground.

news/bellular, and with that, you can get 40% off the Vantage plan, which is what I use. So check them out down below. With that said, let's get into the video.

Denuvo opened a public Discord, which of course was immediately a show, so they made it read-only, and then they re-enabled it, but with far tighter moderation. Then, after all that fun went down, they talked to the press, most notably to Rock Paper Shotgun, who got an interview with a product manager at Denuvo. There's plenty to the interview, so here's the TL;DR: what Denuvo said.

They are no longer willing to let what they say are hostile, toxic, and untrue comments be unchallenged on topics like destroying SSDs and impacting game performance. They say that they know people don't trust them, and that's why they've got a public Discord. They think there are two reasons for their bad rap, right?

Which are: their software actually working, which annoys pirates, so pirates have an incentive to be angry at Denuvo, and that there is no immediate benefit to consumers with Denuvo anti-tamper, unlike Denuvo anti-cheat where no cheaters is going to be a good thing for a customer. In the long term, in the interview, they say that they are in the customer's best interest because they increase revenue, which they say can be the difference between a game getting a sequel greenlit or not. Most notably, they admit that the constant negativity does actually impact their business, which, of course, with Electronic Arts, we can directly see: there's no Denuvo, and that.

. . means the Denuvo will not be paid.

Now, there is a simple question that we all sort of throw back, which is: why don't you release the benchmarks? Well, Denuvo basically responds to that. First, they say that partners can, but they don't.

And Denuvo does have benchmarks, but they can't release those benchmarks without partner approval. Even if they did release benchmarks, nobody would trust them. They then said they would like to have media, like say, Foundry do third-party tests, but that would need to be partner-approved and would need to be on-site in their Austrian facility.

As for performance, well, they recognize problems, but they give examples of basically it not being Denuvo's fault; it being developers implementing Denuvo incorrectly. One example of that they say is Tekken 7. So, their messaging overall here is fairly clear and consistent.

But it still is a bit of a strange interview. As an example, they use an entertainment computer in the University of North Carolina journal to argue that they earn publishers a mean of 15% more revenue and a median of 20% more revenue, which does sound really good. But then they say the survey lacks data on long-term revenue.

You know, it does kind of seem like it's the main point. Anyway, the guy interviewed also says this banging line: "I am a gamer myself, and therefore I know what I'm talking about," which might be true, but it does come off as a bit of a bizarre off-topic defense. At least so far, that is Denuvo's foray into trying to get their image just a little bit repaired.

One thing is for sure, though: a Discord existing and an interview cannot undo a decade of bad PR. So here's why their reputation is so bad and why there's a likely cold business reason why. Quite simply, EA just weren't bothered, but also how EA definitely don't trust you.

Argument one against Denuvo is that it only hurts legitimate players. The idea is that a game is going to be cracked anyway, so what's the point? You're only hurting real customers.

Well, if we're going to say 2016 and then a few years after, yeah, Denuvo would sometimes even be broken before release. But actually, when you look at recent years, like 2023 and 2024, that's actually not really the case anymore. Denuvo does appear to be significantly more resilient than it was in the past.

In 2024, just one game has been bypassed—yeah, one game—and that's Metaphor: ReFantazio. But the reason why it was bypassed is because the demo they released didn't have Denuvo. So, because people had access to the demo, they were able to build a workaround.

Now, looking back at 2023, you do see that more Denuvo games were actually not cracked than were cracked, which is definitely their technology getting more effective. Denuvo essentially used to be anti-piracy hopium or just a bit of a delaying tactic, but nowadays, well, yeah, it has actually got stronger, and that does mean the Denuvo hits piracy a bit harder than it used to. But that's not to say that Denuvo stays forever; as an example, loads of EA single-player games that launched with Denuvo no longer have it.

That's really because most companies seem to find that the anti-piracy measures aren't worth the cost or the effort after the launch window of their game. Basically, Denuvo is not a pay-once thing. I'll get into their payment later on, and it does have ongoing costs.

So, as soon as the expected cost-benefit analysis flips, most companies just choose to no longer use Denuvo. So, argument one, then, that it only impacts legitimate customers—that was very, very true back in 2016 when Denuvo was routinely cracked. Nowadays, at least going by the numbers, it does seem that Denuvo is significantly more effective and that it is less crackable.

That means that, yes, it does actually impact illegitimate use. Argument two, though, is the one that people have really felt, and that is the issue of performance. This is actually where things get more interesting and nuanced than I thought going into this story.

For years, we've all been pretty adamant that Denuvo does impact performance, and that was helped by continual examples of the cracked game running way better than the Denuvo counterpart. But throughout all of this, Denuvo has constantly held the line that they don't impact performance, and that just felt like gaslighting, which, you know, wasn't really fun. It's probably one of the reasons why they ended up being so hated.

What they basically say is that if it does happen, it's not their fault. We can take the examples of Resident Evil Village and Devil May Cry 5—mostly Village, though. So both games had cracked versions that ran significantly better, and Digital Foundry's testing did actually confirm that.

But then Capcom realized that they made mistakes when implementing Denuvo in one of those cases. We specifically know that Village had its Denuvo implementation fixed, and that both games did see performance improvements. Another example is actually Assassin's Creed Origins, where their implementation of VM Protect pushed up CPU usage, which was obviously pretty bad because Origins was a very CPU-bound game in many cases.

Now, things get a bit more interesting with examples from some modders. One of them is Durante, who you may remember—basically, he's the guy that fixed Dark Souls on PC. He found that F15 had increased load times with Denuvo but essentially identical in-game performance.

Another example is Morris Human, who bypassed Denuvo on Hogwarts Legacy, and he posted this. I'm just going to read it verbatim so I get it across exactly: "Looking at FPS and measuring drops is going to yield either two scenarios. " There are no drops at all, which means everything is super optimized; or there are FPS drops.

In the case of FPS drops, the results are going to be inconclusive, as it could mean Denuvo is bad, the game is bad, or the most likely scenario: my hooks are bad. Essentially, what he would be doing would be causing some of those FPS drops. So, at the very least, that was interesting to hear from him.

I think the ultimate solution here is third-party testing actually happening. That's going to be challenging because you're going to have to pay Digital Foundry or some other group to actually go out there, and then you're going to have to have partners actually greenlight that, which could be a challenging thing. Ultimately, though, third-party benchmarks are the only gold standard way to actually have this happen.

And I suppose if there are often problems where a company will implement Denuvo in a way that will have a performance impact, one thing is that Denuvo says they're trying to make that better and easier. But the other point is: how about we don't ram things into games just before launch when there clearly isn't enough time to implement them well? That’s something which, again, is not how a lot of AAA games end up being released.

Ultimately, this is just complex benchmarking; highly technical stuff is very challenging. And it's not just Denuvo; it's Denuvo and how a developer implements it, which basically means there are loads of different variables. Personally, I think there are likely times when Denuvo has been a problem; I think there are likely times when the developer's implementation has been a problem.

It's probably both. But there's a reason why, at the end of the day, that doesn't matter as far as an end user is concerned: Denuvo was in their game, and their game ran worse. Does it matter if it's the software engineers at a developer implementing Denuvo and causing the problem, or Denuvo itself?

Not really, because, at the end of the day, all you know is you bought a game legally, it has anti-piracy protection that doesn't have an immediate positive benefit to you, and your game now runs worse—pretty obviously bad. Now, there is then the other problem: Denuvo games must be connected to the internet. This is the one that I think is actually the worst.

If you're not connected to the internet, then, um, yeah, they'll just refuse to activate. This is obviously anti-consumer and anti-DRM, and of course, DRM generally is about balancing us against piracy. And when I say "us," I mean legitimate people who want to buy a game normally.

As much as Gabe says piracy is a service problem, that is true, but Steam does not sell DRM-free games. You do accept Steam DRM on every single Steam game, and while offline play is an option there, it is not one without considerations and things you've got to do on your account. With Denuvo and the always-online component, there are examples of this impacting players.

I mean, number one: just not having internet, right? So there is that. But also, for six hours in May of this year, you couldn't play Persona 5 Royale; Denuvo denied it was them, but players found a workaround by basically launching the game with no internet services, and they were able to get it working.

Now, 2021 had a similar issue; a Denuvo domain expired, which meant that people lost access to Tomb Raider, Mortal Kombat 11, Total War: Warhammer, Guardians of the Galaxy, and Planet Zoo. At least in that case, Denuvo did confirm that the problem was them. Now, at least for consumers, there is a bright side here, and that is that games actually keep Denuvo around for less and less long these days.

But I think for most of us, it's the principle that matters: we buy a game legitimately that has anti-tamper software in it, where that restricts our ability to access the game and at least causes a risk of decreased performance. All of this leaves Denuvo in a strange position; they cannot prove they don't cause the performance problems. I mean, I'm sure they would love to; they certainly seem to say they have benchmarks that prove it, but their partners, they say, won't let them release the benchmarks.

So the best they can do is “trust me, bro,” which obviously people are not going to buy. Denuvo has lost the narrative, so now the best they can do is try to argue that the problem they solve matters. That's why they said that game margins are getting tighter and that they can increase revenue by a median of 20%.

But it is that direct value of your launch window sales that Denuvo exists to protect; that's why it costs money. There are two sources on how much Denuvo costs. First, there are Krych leaks from a few years back.

Those suggested €140,000 for 12 months of protection, €2,000 a month after that, an extra €60,000 for games with over 500,000 unique activations in 30 days, then 40 cents per activation on WE Game (which is a Chinese platform), and 10K per storefront. So there’s that from a few years ago. But then there's this, which is an AWS Marketplace listing.

Now, it says that for a one-month contract, it's $25,000 USD—right, one month, one game—but the bit that's interesting is plus 50 cents per activation. Now, let’s say that your game gets 1 million activations; that's going to be $500,000 plus the $25,000, and that means we've got potentially millions of dollars on the line. So yes, looking at the pricing.

. . No doubt, EA will want to save money, but very notably, whenever Mike Gamble said, "We trust you," that's not fully true.

Now, I'm sure Mike does actually trust people, but EA overall, less so. You see, Vilar had no pre-install period on PC. That means they are happy to say that they trust players, and they want to use a lack of Denuvo as a marketing beat for their game.

But that doesn't mean that they're going to let their code get into the wild early to give crackers a head start. EA are just gambling that the lack of Denuvo ends up earning them more in sales and saving them enough money to make up for how cracking the game will technically be a bit easier. However, they want to sort of hedge themselves—protect themselves against that—by having no preload period.

That's really what's going on. Now, if we're going to take this to a larger place, the question will be: will this pay off for Electronic Arts? Because if it does pay off, and they basically realize that they didn't really need Denuvo that much anyway, then Denuvo is really, really going to be in a pickle.

If a big AAA company with a big new, expensive game decides that even they don't need Denuvo, well, you probably don't need Denuvo either. For Denuvo, that's going to be a problem. And speaking of problems, there's the Concord story.

It's not just the developer being gone; it's how much Sony basically ended up holding the bag and the degree to which it's an almost unprecedented level of failure in our industry. Check it out next.