The Jobs AI Is Coming For First

Why the Adult Industry Is a Special Case

Most industries adopt AI gradually, weighing quality tradeoffs against cost savings. The adult industry is different, for a few structural reasons:

The cost structure favors AI enormously.

Producing content with human performers involves booking talent, studio time, production crews, editing, and — critically — extensive legal and compliance overhead (age verification, consent documentation, health protocols). AI-generated content strips out nearly all of this. A synthetic video can be generated at a marginal cost approaching zero once the underlying model exists.

Personalization is the product.

Unlike most media, adult content has a business model that rewards infinite customization — different scenarios, appearances, and preferences for different paying customers. This is precisely the kind of task-space generative AI is suited for: not just replacing performers, but generating bespoke content on demand, something a finite pool of human performers physically cannot do.

Anonymity and scale cut both ways.

Platforms can generate and distribute synthetic content faster than they could ever coordinate real shoots, and can do so without the reputational and consent liabilities that come with real performers — though this same anonymity is fueling a wave of non-consensual deepfake content, which is a serious harm in its own right and a growing area of legal and regulatory action.

The technology is already commercially ready.

Unlike, say, fully autonomous vehicles, AI-generated imagery and video of this kind doesn't need to clear a high real-world safety bar. It just needs to be visually convincing — a bar that generative video and image models have already cleared for many use cases.

What This Means for Performers

The economic incentive described above doesn't mean human performers disappear overnight — but it does reshape the industry in a few predictable ways:

• Downward pressure on rates, as AI-generated content becomes a cheaper substitute at the margins, especially for lower-budget or niche content
• A shift in what "value" a human performer offers — increasingly tied to brand, personality, fan relationships, and platforms like OnlyFans where the performer is the product, not just the content
• Legal battles over likeness and consent intensifying, as performers (and non-performers) contend with unauthorized use of their image in synthetic content
• New leverage points for performers who own their platforms and audience relationships directly, since AI can copy an aesthetic but not an existing fan relationship.

The Other Five Occupations

The same underlying pattern — AI doing cheaper, faster, or infinitely scalable versions of previously human-only work — shows up across several other fields, though for different reasons.

Catalog and e-commerce models.

Retailers can now generate photorealistic images of a garment on a model of any body type, skin tone, or age from a single flat product photo. Fast-fashion and e-commerce brands, which need enormous volumes of images quickly, are adopting this fastest — and that's exactly the segment that historically employed the most catalog models.

Photo editors.

Generative fill, AI-assisted retouching, and automated color correction have absorbed much of the manual editing work that used to require a trained eye. Increasingly, clients skip photography and editing entirely, generating images from scratch.

Draftsmen.

AI-assisted CAD tools can now generate technical drawings from natural-language descriptions or rough sketches, and engineers are going directly from concept to drawing without an intermediary drafting step.

Legal research.

AI research tools can search case law, summarize precedent, and draft memos in a fraction of the time it takes a human researcher, and billing pressure from clients is accelerating adoption of these tools over traditional paralegal-heavy research teams.

Data analysts.

Plain-language AI copilots can now write queries, build dashboards, and flag anomalies directly, cutting into the routine reporting work that used to make up a large share of entry-level analyst roles.

The Common Thread

Across all six occupations, the jobs most exposed share a few traits: the work is visual or document-based, it's repeatable at scale, and the "value" it delivers can be described well enough in a prompt for a model to approximate it. The roles that survive longest tend to involve relationship-building, judgment under ambiguity, or accountability that can't be outsourced to a model — whether that's a data scientist who understands business context, a lawyer making a strategic call, or a performer whose value lies in a direct relationship with an audience.

The adult industry is simply the sharpest edge of this trend — not because the technology there is more advanced, but because the economic incentive to deploy it is unusually strong.