Stretching Too Far –Claims & AI

In this month’s column we take a look at the challenges faced with the mushrooming of AI in the industry, and the impact on claims compliance, especially ingredients.

In last months column I closed with a quote from Thomas Aquinas that, “Truth is conformity of the mind to Reality.” The problem we have today though, rather than back in the 13th century, is that Reality for some is actually questionable. Many people live and breathe their lives through the internet and social media. The flesh and blood sitting or walking beside them isn’t considered in the same way as it is online. Are we really living the horrors foretold by Orwell and others? Will AI be humanities next warning?

In the cosmetic industry, when I hear discussions about the pros and cons of AI, it always reminds me of those “skin stretching exercises” displayed in all their blood and gore in that famous scene from the Star Trek film Insurrection. So where are we going with AI, and why do we need to display broad shoulders and common sense if we are going to make it work for us, rather than dig the holes we find ourselves in, even deeper? We are seeing the rise of many consumer orientated platforms which screen out toxic substances and claims ¹, and some even claim themselves to review claims and judge if those claiming are valid? How can they do this when they have no access to product PIF’s, and even when brands put up data on their websites, it is not the whole story?

Artificial intelligence is a somewhat ambiguous term since it refers to different technologies and methods used to compute outcomes, decisions, or recommendations. A personal recent example, in which I ended up on the floor laughing, was that an insurance company I was dealing with at the time, uses AI to determine the exactly when I/we would die! As a Catholic, I firmly believe that only God can answer this, not a computer! One thing is certain though, AI is a shiny marketing tool, and our industries marketing creatives are insatiably creative! Our industry’s press, B2B and B2C, are hyped up so much that it may undergo those skin stretching exercises I have just mentioned. AI has become a trendy new claim, and the regulators are “waiting” for you. If you are in the USA, the FTC² are already gunned and prepped.

A recent Twitter post highlighted the report that 40% of green claims 3 have no evidence whatsoever to them back up. As we all know (even those who won’t admit to it), the internet and social media, are full of misinformation/disinformation. Even Google search has got in on the act too, spouting false definitions of cosmetic ingredients and product claims. So how do brands and platforms go about generating claims, or deciding a claim is valid, when the information their flawed AI system is plugged into, is flawed before they start? A dilemma indeed!

When you are using AI to support your product and its “claims”, are you sure that you are not making claims that are just not true? What risks are you taking to do this – you can’t blame the developer of the technology, or can you? Using AI which at present denigrates permitted ingredients (so is non complaint with the EU claims criteria) is a key issue.The regulators and inspectors are going to be very busy, so at least their jobs are safe for now.

And for the Trekkies among you, even Star Treks Data did not want to be just AI, it preferred to become human if “he” could, despite its flaws.

1 https://www.byrdie.com/beauty-ingredient-websites-5080754 (web pages highlighted therein)
2 https://www.ftc.gov/business-guidance/blog/2021/04/aiming-truth-fairnessequity-
your-companys-use-ai
3 https://www.gov.uk/government/news/global-sweep-finds-40-of-firms-greenclaims-
could-be-misleading