← All beauty signals
Governance

The Rendered Result

A before-and-after photo counts as an efficacy claim. Regulators now say a generated “after” image answers to the same doctrine — labelling it AI-made doesn’t excuse an unachieved result.

What is it?

A generated ‘after’ photo lands under the same rule as a real one: UK and US regulators already treat a before-and-after image as an efficacy claim. Guidance now says an AI image gets no exemption, and labelling it AI-made does not excuse a result nobody measured.

Why now?

Generative tools made a plausible ‘after’ image trivial to produce, just as regulators confirmed the old before-and-after doctrine survives AI intact. A parallel EU labelling duty arrives on a separate track. Missing still is a ruling testing an honestly-labelled image that still shows an unreal result.

Who’s building it?

Brand marketing and legal teams, plus compliance reviewers now treating a generated before-and-after like any written claim.

Compliance teamsBrand legal counselAI image generatorsAd regulatorsClaims reviewers
The full read
Why now, who is building it, where the market goes, and what stays unsettled.
Get weekly →
01 · The market
Who’s in it
Brand marketing teams — commissioning generated after-images before any ruling.
Regulators — the ASA and a parallel EU statute, both reaching the same image.
Shoppers — left guessing whether a labelled AI image shows a real outcome.
What’s at stake
Whether an honestly labelled AI ‘after’ image can still count as proof, once no photograph stands behind it.
A compliance layer is forming ahead of one ruling.
No brand wants to be the first test case.
What’s unsettled
Does a disclosure label change the impression a shopper forms?
Will the EU’s synthetic-labelling duty ever reach efficacy claims?
Whether a beauty brand will be the first tested on a generated result.
02 · How the demand matureda progression, not a timeline
Photographic proof
a real, dated photo was once the only evidence available.
Generated plausibility
a rendered image now supplies a result no camera ever recorded.
Guidance without adjudication
the ASA says the doctrine reaches AI, untested against a labelled, false one.
Owning the exposure
a brand publishing first carries a risk no regulator has priced yet.
03 · Where the market goesthe forks, not a prediction
→A ruling could land either way — disclosure never cures a false result, or a labelled render sits outside the old doctrine.
→Marketing and legal teams start evidencing a generated image the way they evidence a written claim.
→The EU’s labelling duty tightens around synthetic imagery on a track separate from efficacy law entirely.
→A brand could simply wait, and let a rival draw the first ruling.
Underneath every fork sits one unresolved question: whether a disclosed, honest label can ever excuse a picture of a result nobody measured.
04 · The Image-Trust deskadjacent signals, same doctrine
The Invisible Filter
Disclosure laws assume viewers can’t tell.
Norway passed its ad-alteration disclosure law 72 votes to 15.
It's Nice That
The Previewed Procedure
A surgeon calls the preview a surface guess.
Dermatologists coined ‘Snapchat dysmorphia’ for filtered-selfie patients in 2018.
ScienceDaily
The Unretouched Promise
A brand extends a no-retouching pledge to AI.
Aerie says its no-AI-bodies rule extends its 2014 no-retouching pledge.
Marketing Brew
05 · The Wire
01
Disclosure doesn’t cure a misleading AI-generated image.
ASA (CAP) ↗
02
Before/after has been an efficacy claim for over a decade.
ASA (CAP) ↗
03
UK regulator treated sped-up before-and-after clips as an efficacy claim
Advertising Standards Authority ↗
04
UK ad regulator: advertisers answer for what AI tools output
Advertising Standards Authority ↗
05
EU requires clear AI-content labels in key cases from 2 August 2026
European Commission ↗
The panel on this signal
They agreed on
–The doctrine is real and old — the open question is whether a labelled AI image escapes it.
–Disclosure of AI use never cures a false claim.
–No adjudicated case has yet tested an honestly labelled AI ‘after’ image directly.
The claims-substantiation scientist disagreed on
–A single generated image can’t be substantiated at all — the real gap predates AI, never fixed for photographs.
Know more about the method →
✦How this signal was built
01 · Panel
A Delphi panel of 4 AI analysts, each reasoning independently.
→
02 · Debate
They challenge and revise each other’s reads over two rounds.
→
03 · Evidence
Every claim is tied to a cited, linked source on The Wire.
→
04 · Synthesis
A moderator writes it into one read, disagreement and all.