By Claire Piech and Lauren Erdelyi
“When I see AI content, that usually turns me right off a company, even one I’ve done business with before.”
— Anonymous
It’s been three years since AI tools hit the mainstream with OpenAI’s launch of ChatGPT. The public’s reactions have shifted dramatically in this period, running the gamut of shock, awe, fear, dismissiveness, speculation and more. These emotions are central to understanding AI’s role in communications, since public reaction has a direct impact on success or failure.
Now in the third year of mainstream AI, we can say with confidence two clear trends are emerging for communications.
1. People can spot AI content, and they don’t necessarily like it
The allure of AI is obvious: low cost, scaled output, minimal effort. What is becoming increasingly apparent, though, is that the public can identify AI-generated content.
This was not the case in 2023. While there were telltale signs, like the over-reliance on the em dash (see: “Is the Em Dash Dead?”), for the most part we were all so fresh that the predictive language sequences were mesmerizing.
Now everyone has ways of spotting AI. For some, it’s words like “delve” when a simpler one would do; for others, it’s the “it’s not X, it’s Y” construction. It can be hard to describe. Paragraphs with a sequence of sentences all the same length. Extremely emotional prose that follows a similar pattern of short sentences at the end.
It turns out people don’t really like seeing it. The lead quote at the top of this post was from a professional forum in which Magnetic is involved. The Globe and Mail likewise ran this recent piece. New terms have entered our vocabulary like “AI slop” and “brain rot”. Even the communications guidelines for International Women’s Day explicitly mentioned avoiding AI-generated content for social media posts.
At Magnetic, we believe the implications of what this means for science communications are clear: AI-generated content runs the risk of alienating audiences and eroding trust. In most cases we recommend companies avoid using AI to create their communications. If a company does choose to use these tools, they must label so explicitly.
This does not mean you should not use AI in your work. AI is a key part of business in 2026 and allows all of us to achieve outcomes not possible before. What it means is you should not use AI to create your communications from the ground up and expect strong, long-term results.
2. Message consistency has never mattered more
In parallel, another communications trend has emerged: key messages are more important than ever.
In communications, key messages are the most essential points you want your audience to remember. They are a cornerstone of best practice. Your key messages should be applied consistently across all your channels, from your website, to your pitch deck, newsletter, social media, press release, blog post and more.
Now that large language models are able to search all your past digital communications in less than five seconds, they are also able to spot, in a way not possible before, where your messaging is inconsistent. For example, if your company describes its mission one way on your website and another in your press releases, an AI bot will know. Any discrepancy means AI will either rate your company unfavourably or not include you in a search altogether.
The takeaway is that you need to lean into communications best practices and ensure your messaging is consistent across all your channels. You could also consider using AI tools to spot deviations.
Looking ahead
AI is still in its infancy, and the technology is evolving fast. It would be impossible to predict what the future will bring for science communications. We will continue to watch the evolution closely so we are always providing the best communications advice.