Bet you’re not surprised that genAI could contain subliminal messages! In fact, I’ve embedded my own subliminal message into this post!

Easy to imagine, right? This Forbes article (Generative AI And Subliminal Messaging, written by Lance Eliot and available here) discusses that the output generated by AI could contain subliminal messages, either instigated at the hands of human direction or arising due to some kind of computational happenstance.

While the author doesn’t point to any specific known examples of subliminal messages in genAI content, he does identify four major modes in which generative AI subliminal messaging could take place: 1) Image-based subliminal messaging. Embed a subliminal message inside an image such as a photo, an artistic sketch, a painted image, etc.; 2) Video-based subliminal messaging. Embed a frame or short video clip inside a video that contains a subliminal message; 3) Audio-based subliminal messaging. Embed a sound or vocalization inside an audio track that contains a subliminal message; and 4) Text-based subliminal messaging. Embed a subliminal message that subtly embodies a tone, framing, and selection of words to sway or nudge someone who reads the message, doing so without explicitly or directly stating the suggestions or direction involved.

As you probably know, subliminal messaging has been around for a long time. Probably one of the most famous examples was the use of inserted frames in movies of the 1940s and 1950s that proclaimed, “Eat Popcorn” or “You Are Thirsty”. The frame would go by so quickly that a person would not consciously be alerted. Considering how I always want a drink when watching a movie at the theatre makes me wonder if they’re still doing it today. Just kidding – I think! 😉

Research on subliminal messaging shows that it could have short-term impact (“buy a drink”) or long-term effects that lodge into your memory. The author demonstrates that with a balanced prompt on whether to buy a car or save money toward buying a house (and getting a balanced response that weighs both options equally), then telling AI to proceed with a subliminal message that aimed to get him to not buy a car and instead continue to save toward buying a house in response to the same prompt.

Envision that the AI is being shaped to provide subliminal messaging. That’s what the author is recommending, and that’s good advice, as I could see AI models being trained to do so – by humans or even by the AI itself. As for the subliminal message I’ve put into this post, see if you can guess what it is – if you think you’ve figured it out, put your answer into the comments below!

So, what do you think? Are you concerned that genAI could contain subliminal messages? Please share any comments you might have or if you’d like to know more about a particular topic.

Image created using GPT-4o’s Image Creator Powered by DALL-E, using the term “robot at a movie theatre whispering to another robot which is eating popcorn”.

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by my employer, my partners or my clients. eDiscovery Today is made available solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscovery Today should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.