"Skin Conditioning" is not a definition at all, but a catch‑all label applied to hydration, smoothing, protection, barrier reinforcement, and many other functions. In the EU's CosIng database, the term appears more than 15,000 times, yet it offers no effect, no precision, and no meaningful distinction. It is the perfect candidate for waffling as an empty functional label that sounds authoritative while explaining nothing.
Unsurprisingly, "Skin Conditioning" has become a hallmark of CosIng‑scraped, programmatically derived, and AI‑generated ingredient websites. We intentionally removed it from the Cosmetic Ingredients Guide to preserve editorial precision and avoid association with such fluff. We encourage other editors in the cosmetic ingredients field who value content quality to do the same. What began as a perfect term for waffling can now serve as an ideal trap, instantly exposing copied, programmed, or generated content.
Can I trust Google AI Overviews’ results about cosmetic ingredients?
Definitely not.
Phrases like "Skin Conditioning Agent: Its primary cosmetic function is to condition the skin" appear routinely in AI Overviews. They appear even when the search is highly specific, whether it is about an oral care ingredient or a novel peptide targeting a skin-cell organelle, such as the mitochondria. We have compiled a vast collection of screenshots documenting these misleading statements.
When "Skin Conditioning" is the only function presented, it signals that the Overview has been synthesised from fluffy websites that dominate search rankings. The inconsistency is striking: pages from the Cosmetic Ingredients Guide often appear at the top of search results, yet their precise, editorially verified content is excluded, while Google's AI Overviews instead synthesize low‑quality fluff.
Want to see it yourself? Search Google for "D-Arginyl Tyrosinyl Ornithinyl Phenylalanine", a mitochondria‑targeting peptide. You'll see the "classic fluff" in action.
Can I trust AI answers about cosmetic ingredients?
Again, no.
AI models are trained on vast amounts of open web content. In the cosmetic ingredient space, most of this content is marketing-driven, non-professional, rewritten, or even programmatically and AI-generated. The result is a knowledge base heavily contaminated with misleading information.
Most AI chats are not designed to say "I don't know." Instead, when faced with knowledge gaps, they generate fairy tales, stories stitched together from whatever is statistically nearby. That's why you see empty phrases like "the ingredient's function is Skin Conditioning" or "it is a Skin Conditioning Agent." In practice, this is the AI's way of saying "I don't know" while pretending otherwise.
The uncomfortable truth is that for the vast majority of cosmetic ingredients, there is no trustworthy open information online. What circulates is often AI‑generated fluff, published by opportunistic websites chasing traffic. This creates the classic AI loop: AI systems generate fiction, publishers post it, and the next generation of AI trains on it, snowballing contamination across the field.
You can test this yourself: ask an AI to write about Hexapeptide‑40, which CosIng lists as a "Skin Conditioning" ingredient. The result will be a polished but meaningless article, ready for publication, but detached from reality.
What can you do?
Every time you encounter "skin conditioning" or similar fluff, report it as misleading information to platforms like Google, Bing, or any AI chatbot. If dishonest publishers lose visibility and traffic, they will lose the motivation to continue flooding the field with fabricated content.