Location: Da Nang, Vietnam — desk, late evening
Context: Reviewing January 2026 search behavior shifts (GEO, experience signals)
Status: Observational
Short version:
AI images are not banned.
Google is not “penalizing” sites for using them.
But perfect images are increasingly a liability for experience-led brands.
With the latest Google Authenticity core update, the bias has shifted—from polish to presence.
1. Imperfection Is Now a Trust Signal
Google’s recent updates (Helpful Content, Hidden Gems, Experience signals) are trying to filter out mass-produced sameness, not tools.
The AI problem isn’t that it’s artificial.
It’s that it’s too clean.
- Perfect lighting.
- Perfect symmetry.
- Perfect composition.
In travel, health, and personal brands, that visual language now reads as:
“Generic. Stock. Content farm.”
Imperfect photos do something AI still can’t fake well:
They prove you were there.
- A grainy gym mirror shot.
- A badly framed street stall.
- Uneven lighting.
- Motion blur.
These are not flaws anymore.
They are evidence.
SEO consequence (important):
There is no technical penalty.
The effect is behavioral.
Pages that feel real keep users longer.
Pages that feel synthetic lose trust faster.
Google reads that outcome – not your image source.
2. The Personal Brand Split (Your Actual Advantage)
For entity-driven sites, polish still works.
For you as a visible, named human credibility now flows through experience proof.
Where Real (Imperfect) Photos Are Non-Negotiable
Proof of location
If you write about Da Nang, Chiang Mai, or Koh Samui, use your photos, even if they look “ugly”.
An AI “Vietnam coffee shop” image quietly kills trust.
Proof of results (OldGuysFit)
Your real body > any generated “fit older man.”
Lighting doesn’t matter. Presence does.
Proof of ownership
Supplements, gear, and microphones show them in your hands.
Vision systems are getting better at detecting possession.
These images anchor Experience the new weighting in E-E-A-T.
3. Where AI Images Still Make Sense (and Are Still Strong)
AI isn’t the enemy. Misuse is.
Good AI use cases:
- Scientific or biological concepts (you can’t photograph magnesium binding)
- Abstract ideas (metaphors, decision forks, system diagrams)
- Background visuals, textures, banners
- Conceptual thumbnails (not claiming real events)
In these cases, AI clarifies – not replaces -experience.
4. The Uncanny Valley Risk (Personal Brand Killer)
The fastest way to collapse trust:
AI-generated humans that are almost real.
Especially:
- Faces
- Bodies
- “You-like” avatars
- People you supposedly met
Rule of thumb:
Never use AI to represent a real person in your story.
One uncanny image can undo ten good paragraphs.
5. What This Quietly Changes About SEO in 2026
This is the part most people miss:
Google isn’t rewarding imperfection.
It’s rewarding believability at scale.
AI made perfect content cheap.
So imperfection became scarce.
Scarcity creates signal.
That’s the shift.
Working Principle (Worth Locking In)
Use AI for ideas.
Use reality for proof.
Or more bluntly:
If the image implies “I was there” → it must be real.
If the image explains an idea → AI is fine.