The FDA’s recent barrage of letters calling out alleged violations of federal pharma marketing rules poses a challenge to drugmakers. With the agency rapidly redefining what is acceptable across dozens of untitled letters, marketers need to quickly learn from the regulatory actions to adapt their campaigns.
Ad agency Real Chemistry is applying artificial intelligence to the task.
Even before the FDA crackdown began, the agency had already built an internal AI assistant to capture observations and understand letters published by the FDA, enhancing its existing feedback loops, Alyson Hagert, president of digital and social media at Real Chemistry, wrote in a blog post Tuesday.
The recent FDA actions prompted Real Chemistry to consider whether it could build on its internal experiment to help pharma companies learn from letters issued by the regulator. That idea resulted in Compliance Compass, a system that monitors FDA enforcement letters and uses AI to identify trends.
Compliance Compass reviews brand creative and assets against the enforcement trends it detects in the FDA’s letters. The idea is to use AI to spot patterns in large amounts of text to help pharma companies preemptively identify and counter new compliance risks created by the FDA’s evolving approach to drug ads.
The scale of the clampdown unleashed by the FDA has given the AI lots of letters to analyze. The agency has so far published 51 untitled letters dated Sept. 9, the day President Donald Trump told the FDA to rein in direct-to-consumer advertising. In comparison, the database of untitled letters features just 32 entries published between 2017 and 2024.
Reviewing the letters published last month shows the FDA chided multiple companies for similar issues. For example, the regulator took issue with before-and-after visual juxtapositions that are common in ads but, in its view, can in some cases imply a level of efficacy that isn’t supported by data.
Other repeat criticisms include the use of eye-catching visuals and background music that could impede understanding of side effects and contraindications.
In theory, AI could cut the time and effort needed to identify such patterns in the FDA’s letters, enabling companies to make timely changes to materials and avoid becoming the next target of an untitled letter.