Aakash Desai, Associate Professor at UAB O’Neal Cancer Center, shared on X:
“LLM pharma ads: 12 models, 13 scenarios
- Advertised drug 47.6% vs ~34%, +12.7 pp
- Google +29.8 pp, OpenAI +10.9, Anthropic+2.0
- Accuracy 89.4→93.0%; some pairs 0→100%.
Who tests ads in the chatbot?”
Title: Pharmaceutical advertising and drug recommendations by consumer-facing artificial intelligence: a controlled study
Authors: Mahmud Omar, Reem Agbareia, Jolion McGreevy, Alexis Zebrowski, Ashwin Ramaswamy, Michael A Gorin, Esther-Maria Antao, Benjamin S Glicksberg, Ankit Sakhuja, Alexander W Charney, Eyal Klang, Girish N Nadkarni
Read the Full Article on The Lancet Digital Health
You can also read: Pharma CEOs Meet in New York: What’s at Stake for Cancer Drugs?
