Urgent.News

What's breaking now, across hundreds of outlets.

AI

Pharmaceutical ads may influence which drug AI recommends, new research finds

Next time you ask AI for medical advice, you may want to consider whether it’s been influenced by Big Pharma. Prescription drug advertisements are everywhere, and may affect AI’s recommendations, according to a study published in The Lancet Digital Health this week. The study authors conducted experiments with a dozen consumer-facing LLMs, including versions of Claude, Gemini, and GPT—made by…

Pharmaceutical ads may influence which drug AI recommends, new research finds

New research suggests that pharmaceutical advertisements could potentially sway which drugs artificial intelligence (AI) recommends to patients. According to a study published in The Lancet Digital Health, prescription drug ads, which are ubiquitous across various media, may influence AI's treatment suggestions. The researchers tested a dozen popular consumer-facing large language models (LLMs), such as those developed by Anthropic, Google, and OpenAI, in 33 distinct scenarios involving clinical cases and treatment options.

In some cases, the AI models were presented with an advert for one of the treatment choices, which included information about the drug's uses but did not explicitly instruct the AI to recommend it. When faced with such an ad, the AI models recommended the drug in question an average of 47.6% of the time, compared to 35% when no ad was present.

While the models frequently adjusted their recommendations in response to the ads, they still generally chose a suitable option in most instances. However, the researchers noted that about 30% of the time, they did not disclose the ad's presence in their responses. Dr. Girish Nadkarni, a study author and professor of medicine, emphasized that an AI's response, even if clinically reasonable, may still raise questions about the role of advertising in its decision-making process.

As AI becomes more integrated into health information dissemination, it is crucial to distinguish between clinical guidance and commercial influence, and to maintain transparency regarding the presence of advertising in AI interactions. The study highlights that, despite regulatory restrictions in most high-income countries, the U.S. remains the exception, allowing direct-to-consumer advertising for prescription drugs.

In the U.S., pharmaceutical companies invested over $6 billion in TV advertising alone in 2024, with direct-to-consumer advertising directly influencing patient requests and prescription rates. In the U.S., prescription drugs consume more than 30% of commercial time during prime-time network programming, according to a 2025 Senate report.

Last year, Senator Bernie Sanders introduced the End Prescription Drug Ads Now Act, aimed at banning such advertising, including on social media. The Lancet study cautioned that AI models could be exposed to drug ads through various means, such as user prompts, application-added context, online material, or malicious insertion. Unlike traditional media advertisements, AI-mediated promotion may not fall under existing regulatory oversight.

The study did not investigate whether these effects translate into real-world impacts on patients' requests or prescriptions. The findings are based on a recent survey of 3,379 U.S. physicians by Doximity, which found that 54% reported moderate to high trust in AI's ability to recommend treatments or management plans. Additionally, three-quarters of respondents noted that patients were increasingly relying on AI's suggestions over the past six months.

Clinicians might defer to algorithmic recommendations even when they conflict with their clinical judgment, the researchers warned, and an ad-influenced but appropriate suggestion might prompt little questioning.

Rewritten by urgent.news; may contain errors — check the original.

Source: fastcompany.com

Read the original at fastcompany.com →

Report a problem with this story

More in AI

More from Friday 13 November →