Copy article

Crimson Desert AI art backlash is a disclosure problem

ended 24. March 2026

The Crimson Desert backlash over AI generated art is being treated as a games culture war. It is actually a trust and disclosure problem that will soon hit every industry using generative tools.

Pearl Abyss says some 2D props were created with experimental generative tools during early iteration, and that the intent was to replace them before release. The fact they shipped anyway is the real lesson. Once AI assisted assets are in the pipeline, they become a governance question, not an artistic one. If a studio cannot reliably label what is generated, what is licensed, and what is human made, then it cannot communicate honestly with players or partners.

This is the same pattern seen in enterprise software. AI starts as a shortcut for speed, then quietly becomes embedded, then gets discovered in production. The backlash is not just about taste. It is about consent, provenance, and whether customers are being told the truth.

The contrarian point: disclosure is not optional PR. It is quality control and risk control. The more AI is used, the more organisations will need auditable supply chains for content.

We'd like your views:

  • Should game studios be required to disclose AI generated assets in a standardised way, and if so, what level of detail?
  • How should platforms and storefronts verify disclosure rather than relying on self reporting?
  • Does undisclosed AI content create a consumer protection issue, or only a reputational one?
  • What internal controls would prevent AI placeholders from shipping by accident?
  • Will stronger disclosure norms protect creativity, or push AI use underground?

4 responses from the Newspage community

Copy all

Star Quote
Copy

Rohit Parmar-Mistry
Founder at Pattrn Data
This is less about a culture war and more about supply chain truth. Once generative tools sit inside the asset pipeline, you need to be able to answer simple questions on demand: what was generated, from which model and prompt, on what training assumptions, and who approved it for use.

Studios do not need to publish every brushstroke, but they do need a standard disclosure that is verifiable. Think of it like credits plus provenance: categories of AI use (concept, texture, 2D props, marketing), whether outputs were materially edited, and whether any third party datasets or vendors were involved.

The bigger fix is internal control. AI placeholders ship when there is no asset register, no gate in build tooling, and no audit trail. Treat AI assisted content as a controlled component: tagging in source control, review checklists, automated scans before release, and a release sign off that someone can stand behind.
Copy

Editor
Mitali Deypurkaystha
AI Governance Strategist & CMO at AISensa
Fused figures, faulty fingers, warped weapons and nonsensical wording turned a fantasy game into an accidental AI audit. It wasn't AI-savvy players uncovering AI through instinct. Even a complete AI novice would have spotted it because the output was visibly substandard. This points to a deeper problem than pipeline checks: why a major studio was relying on tools or outputs that fall so far below today’s baseline. When even casual gamers can see human figures fused into horses creating strange centaur-like glitches, it suggests both poor tool choice as well as absent human review. Yes, the 'excuse' is they were meant to be placeholders, but I'd expect even placeholder images to be of a high level from the likes of Pearl Abyss. The real risk isn’t just undisclosed AI, it’s low-quality AI quietly lowering standards and breaking trust, especially when companies aren’t open about how it’s being used.
Copy

Editor
Dil Gujral
Chief AI Trainer at AI Now Academy
Modern gaming has pivoted from a creative medium to a high-efficiency extraction model.

Driven by predatory monetisation, ranging from manipulative loot boxes to FOMO saturated battle passes.

The industry now prioritises "wallet share" over genuine user experience.

Expecting corporate benevolence in this space is a strategic error.

These systems are designed for extraction, not entertainment.
The real concern is the dilution of human capital.

The immense cognitive bandwidth and problem-solving skills honed in digital environments are currently being harvested for zero real-world ROI.

It is time for a strategic pivot, gamers should rechannel that same mindset toward the "game of life."

And, parents need to do choose the hard option and start saying NO to computer based gaming.
Copy

Colette Mason
AI Ethics Consultant at Clever Clogs AI
"It was only placeholder art" is the AI equivalent of "the dog ate my homework." If your pipeline can't tell placeholder from production, that's not an accident. That's a missing control.

This isn't about games. Every business using generative tools faces the same question: can you trace what's AI made, what's licensed, and what's human? If not, you don't have a creative workflow. You have a liability you haven't found yet.

The backlash isn't the interesting bit. Pearl Abyss meant to replace the AI assets and didn't. Nobody owned the handover between "experimental" and "final." If your production pipeline can't distinguish AI generated from human made from licensed, that's not a creative decision. It's an unaudited risk. Workers and workplaces need labels for these things, or it will never get discussed.

Disclosure isn't PR. It's provenance. And provenance is the only thing standing between generative tools being useful and being a trust wreck across every creative industry at once.