The texture of reality is the new luxury

Paula Ironside
May 7
8 min read

Why the brands winning beauty and wellness in 2026 are using AI where it adds leverage, and refusing it where it erodes trust.

midjorney model — Image created by me in Midjourney

The brands winning right now aren't the ones using AI the most. They're the ones using it where it adds leverage and refusing it where it would erode trust.

That's the whole post. The rest is the working out.

I've been using AI for over three years now. Long enough to have seen both the good and the bad sides, which I want to talk about too. And since my job is to pay attention to what's happening in beauty and wellness, I keep coming across the same split: brands using AI as scaffolding, and brands using it as a substitute. Only one camp is building anything that lasts.

(Mind you, I used Ai as a tool to help me write this blog too. So this isn't a sermon. It's a conversation about how.

Here's the principle that holds across everything below:

If AI shapes your thinking, it helps. If it is the thinking, you've already lost.

If you take nothing else from this post, take that.

What's actually happening to your customer

Before we get into do's and don'ts, you need to understand what your customer is doing right now, because it changes the stakes.

Three things are true at once.

One. Your customer is increasingly asking ChatGPT, Perplexity, and Google AI Overviews for product recommendations instead of searching by brand. A quarter of US internet users will use generative AI for shopping in 2026 (eMarketer). For skincare, ChatGPT now leads Google for concern-driven queries, eczema, acne scars, dark spots (SPATE, 2025). If your brand isn't legible to a language model, you're being skipped before the customer even sees you.

Two. Your customer is also exhausted by AI content. 76% of US adults say it's extremely or very important to be able to tell whether something was made by AI or a human (Pew, 2025). Nearly 70% are concerned AI-generated content will be used to deceive them (Deloitte, 2024). 58% have stopped buying from a brand after discovering misleading product information (Deloitte Beauty Trust Index, 2025). When McDonald's Netherlands ran an AI-generated Christmas ad in late 2024, comments like "ruined my Christmas spirit" and "AI slop" forced them to pull it.

Three. Your customer wants both. 76% of beauty consumers are open to a trusted AI-powered personal shopper (Accenture, 2025). 64% of Gen Z are more likely to buy from brands offering AI-driven solutions, but 61% still want a combination of AI and human consultation (Revieve, 2025).

This is the contradiction you're navigating. AI is unavoidable in discovery. AI is increasingly resented in storytelling. The brands handling this well are the ones who internalised the difference.

When AI is a trap (and what the trap actually costs)

Let me give you an example I see almost daily. You're a small natural cosmetics brand. You can't afford a photographer. You generate images, tell yourself "this looks pretty good," and post.

The viewer clocks it. Not always consciously, but the friction is there. "Is the product real or fake too? Why am I paying a premium for this?" You haven't just made one bad post. You've quietly devalued every claim you've ever made about made with care, real product for real people, timeless formulas.

Three places I'd refuse AI outright:

1. AI influencers and AI "team members." SheerLuxe, the UK fashion publication, introduced an AI team member in 2024 without flagging her as AI. The backlash was loud and lasting, it's now the reference editors and strategists use when they want to say don't do this.

2. Final visual identity for a brand built on human values. If your brand language is care, ritual, ingredients, natural, craft, the over-glossy, suspiciously symmetrical AI aesthetic actively fights you. Worse, it's becoming dated fast. The current AI look will read in 2027 the way 2014 Instagram filters read now. If you build your visual identity on it, you're locking yourself into a moment that's about to expire.

3. Final written copy. AI is trained to make us satisfied. It nods along. It softens. It defaults to "you're so right!" In a feed full of pointless posts every second, a recognisably human voice, opinionated, specific, willing to disagree, is the actual moat.

A note specifically for wellness founders

If you're in supplements, sleep, longevity, mental health, fitness, or anything adjacent to clinical claims, your AI risk is not aesthetic. It's regulatory.

AI-generated copy and chatbots tend to subtly overstate efficacy. They imply scientific support that isn't substantiated. They blur aspiration and measurable performance. The FTC, the MHRA, the EU AI Act, ASA in the UK, all of them are tightening enforcement on health claims. An AI-generated product description that overstates your formula's benefits isn't just a brand problem. It's a fine, a takedown, and potentially a class action.

If you let AI write your wellness copy, you need a human with claims expertise reviewing every line before it ships. Treat AI in this category the way you'd treat an unsupervised intern with a law degree they don't have.

The competitive reality (the part most posts skip)

Here's the honest version. Your competitor is shipping 40 AI-generated posts a week. They're growing. The temptation to match volume is real, and "just don't post for the sake of posting" can feel naive when you're watching someone else win the algorithm.

So let's be specific. Volume strategies tend to win short-term and lose long-term. Performance content, short videos showing real product results, real ingredient sourcing, real founders, converts at roughly 2x the rate of generic aspirational content (HatchEcom, 2025). 70% of beauty shoppers are now influenced by online education, reviews, and social proof rather than pure advertising (McKinsey State of Beauty, 2025). The brands that look like they're losing to the volume players this quarter often look like the only credible ones in eighteen months.

The real choice isn't AI volume vs no posts. It's AI volume vs fewer pieces of work that compound. One photoshoot every six weeks, plus consistent BTS, plus three to five nano-influencer relationships, plus a strong written voice across two channels, that beats forty AI carousels for any brand whose product depends on trust. Almost every time.

What indie founders are actually doing well

I want to stay grounded here. The L'Oréal and Estée Lauder examples that get cited in every AI-in-beauty article aren't relevant for most people reading this, they have engineering teams and seven-figure tooling budgets. So let's look at what founders working at our scale are actually doing.

Régime des Fleurs, the indie fragrance brand, is treating AI as inevitable infrastructure rather than a creative shortcut. "AI is happening whether we like it or not. No business wants to be left in the dust," founder Alia Raza told Beauty Independent earlier this year. Note what she didn't say, she didn't say AI is generating the brand's voice, visuals, or storytelling. The brand's aesthetic universe is still built by humans. AI is the back-office layer.

Stephanie Scott-Bradshaw of First and Last PR put it cleanly: "We integrate ChatGPT as a strategic enabler rather than a replacement for human creativity. We use it for deep research and data analysis, surfacing insights that inform our campaigns, as well as for smaller yet time-consuming tasks like organising reports, note-taking, and mapping key initiative timelines." Research, organisation, repurposing across platforms while maintaining a consistent and authentic voice. That's the line.

Renée Cosmetics, an indie Indian makeup brand, took a different angle, they used AI in customer infrastructure rather than content. They implemented an AI-driven WhatsApp cart-recovery tool and reported a 63% conversion rate on abandoned carts. Boring on the surface, transformative for the business. AI in operations, not in storytelling.

Erly, a science-driven indie skincare brand co-founded by a dermatologist, is leaning into the regulatory side of the wellness conversation, using AI tools to translate clinical and ingredient information into something consumers can actually understand, while keeping clinical authority in human hands. AI as translator, not author.

The pattern across these is what you'd expect: AI is helping with research, customer infrastructure, ops, and translation. None of these brands are letting AI generate their visual identity, brand voice, or POV.

There's a public refusal play worth knowing about too. Dove's 2024 "The Code" campaign committed to never using AI to represent women, and released a 72-page Real Beauty Prompt Playbook for other brands. The result was 500M+ organic views, 94% positive sentiment, and the biggest PR coverage in Dove's history. Worth saying: Dove is owned by Unilever, who also owns brands using AI heavily in marketing, so the position is selective, not absolute. But the strategic lesson is real, and it's accessible to indie brands too. Aerie pulled the same move with their "no AI bodies, no AI models" pledge, building on a decade-old "no retouching" promise, and saw measurable engagement lift.

For a small founder-led brand, this is actually the most accessible play of all. You can't out-spend the giants on AI infrastructure. You can absolutely out-real them. And in 2026, that's a real competitive position.

How I use AI in my own practice

I'm not writing this as a journalist. I'm writing as a strategist who lives this. So here's my actual line.

Where AI earns its place in my work:

Research and cultural pattern-finding. Synthesising hundreds of articles, trend signals, founder interviews. AI helps me see patterns faster across the noise. But the pattern I notice, the one that becomes a strategy, is mine.
Writing scaffolding. Outlines. Reordering. Editing. Stress-testing my logic.
Client-facing operational tasks. Proposal structures, deck reformatting, summary documents. The plumbing.
Structure and systems so nothing gets lost or forgotten, especially when dealing with hundreds of files and exports.
Quick fixes like retouch or copy correction

Where I refuse it, even when it would be faster:

Creative direction and art direction. Every shoot, every visual, every reference board. The eye is mine. Brand positioning, visual decisions, taste, these can't be delegated to a model trained on the average.
Visuals published as "real" without "made with ai" mention
Anything client-facing that represents the brand's POV. The Substack, LinkedIn pieces, client communications, written by the client team or myself.

The principle underneath all of it: if AI shapes my thinking, it helps. If it is the thinking, no.

That's the line. It works because it's about ownership of the work that defines my brand, not about avoiding a tool.

The decision framework (screenshot this if you want)

Three questions to ask before any AI-assisted post or asset ships:

Does this feel real?
Does this build trust?
If a customer found out exactly how it was made, would they feel respected, or tricked?

If the honest answer to any is no, the post isn't worth what it costs you.

What this is really about

The deepest version of this conversation isn't about AI. It's about brand sovereignty.

AI is the symptom. The real story is that founders are losing control over how their brands are perceived, recommended, and remembered. Discovery is moving into AI assistants you don't own. Aesthetics are converging into a synthetic average you didn't choose. Storytelling is being commoditised by tools your competitors also have.

The defensible position is the one no model can replicate: a specific point of view, executed with specific taste, in a voice that's recognisably one human's. AI can help you produce more of that. It cannot generate it for you.

The brands that understand this in 2026 will be the ones still here in 2030.