Yes and no. That’s not a hedge. It’s the most accurate answer the data supports.
Some design work is being replaced right now. Some design work is in higher demand than ever. The difference is not about seniority or experience in the traditional sense. It’s about what kind of design work you’re talking about, and how clearly you can draw that line.
This post gives you the honest version: what the actual numbers show, what AI replaces and what it doesn’t, and what it means if you’re a founder making decisions about design.
The numbers that worry people
Graphic designer job postings dropped 33% in 2025. That’s real. Oxford Martin School puts graphic designers at 86% on their automation risk scale. Anthropic’s 2026 Economic Index shows 37% observed AI exposure for design roles. These are not made-up concerns.
If you make generic social media graphics, resize assets for different formats, create templated marketing materials, and do production work that follows a clear brief without strategic input, AI tools can do a meaningful portion of that now, and they’ll do more of it next year.
So yes: commodity graphic design is being automated. That part is true.
The numbers they don’t lead with
74% of businesses say AI has either not changed their need for design professionals, increased the demand, or only replaced simple tasks.
Only 31% of designers currently use AI for core work, according to Figma’s 2025 AI Report, compared to 59% of developers. The profession hasn’t even fully adopted the tools yet, and demand is holding.
UX design roles are projected to grow 16% through 2026. Graphic design 2%. The more a design role involves judgment and strategic thinking, the more it’s growing. The more it involves production execution, the more it’s at risk.
Designers with AI skills earn 56% more than peers who don’t use AI tools. Design job listings mentioning AI skills went from 3% in 2023 to 32% today. Autodesk’s AI Jobs Report found that design has overtaken technical expertise as the most in-demand skill in AI-related job postings.
AI is not reducing demand for designers. It’s increasing demand for a specific kind of designer, and that designer earns more.
What AI actually replaces
AI replaces or significantly reduces: production work that follows clear specifications, asset resizing and reformatting, first-draft generation for standard layouts, color scheme variations, icon sets from descriptions, presentation template population, and certain wireframing tasks where the structure is already defined.
This is real work that real designers spend real hours on. A substantial chunk of a junior or mid-level graphic designer’s week falls into this category. That’s why the 33% drop in job postings is concentrated in those roles.
What AI doesn’t replace
Strategic creative direction. The judgment call about why a landing page isn’t converting, even though it looks fine. The decision to simplify an onboarding flow that users are quietly abandoning. The understanding of what a specific user cohort does and doesn’t notice.
AI doesn’t know your users. It can’t sit in a user interview and notice that three different people described the same problem three different ways. It can’t look at your analytics and spot that drop-off is caused by a loading state that feels like a crash.
It can’t tell you that your pricing page is structured for the way your product team thinks about tiers, not the way your buyer makes a decision.
These are judgment problems. The research on automation risk consistently identifies strategic creative direction, brand identity, and UX research as the most AI-resistant skills in design.
The designer who earns more because of AI
A senior designer who uses Figma’s AI agent to iterate 4 versions of a component in the time it used to take to produce one, then applies strategic judgment to pick the right direction, is more productive and more valuable than they were a year ago.
A designer who uses vibe designing tools to generate a first-draft layout in 20 minutes, then refines it with actual knowledge of the user and the conversion goal, compresses what used to be a two-day process into a single session.
Designers with AI skills earn 56% more than peers without them. That’s not because AI is doing their job. It’s because AI is multiplying their output, and clients pay for output.
What this means for founders
For marketing asset production, social content, and templated work, AI tools are genuinely capable now. If that’s most of your design need, you don’t necessarily need a designer for it.
For product UI/UX, landing pages tied to conversion goals, onboarding flows, and any design that has to understand your specific users, you still need a designer with judgment. AI tools today don’t make those decisions well without senior oversight.
The trap founders fall into: assuming that because AI can generate something that looks like a landing page, it’s generating one that converts for their product. The first is a visual production task. The second is a strategy problem with a visual output.
We see this at DesignShare. Companies come to us after trying AI tools for product design and finding the output looks fine but doesn’t perform. The design is technically correct. The thinking behind it is missing.
