Unregulated health and AI apps for substance use reduction: A critical review
In a recent commentary published in the Journal of the American Medical Association, researchers from Rutgers Health, Harvard University, and the University of Pittsburgh shed light on the impact of unregulated mobile health and generative artificial intelligence (AI) applications in the field of substance use reduction.
The commentary emphasizes the urgent need for greater oversight of new and untested technologies, particularly in the context of mobile health and generative AI. Jon-Patrick Allem, a prominent figure in the Rutgers Institute for Nicotine and Tobacco Studies and an associate professor at the Rutgers School of Public Health, underscores the importance of public marketplaces adhering to stricter rules and regulations to manage these innovative tools effectively.
The core issue lies in the potential for misinformation and the lack of transparency surrounding these apps. While some mobile health apps have shown promise in helping individuals reduce substance use, particularly in controlled studies, their real-world impact is often limited. App stores, driven by ad revenue, may prioritize apps that generate income over those backed by scientific evidence, leading to a proliferation of untested or misleading products.
The Substance Use Reduction App Dilemma:
Research indicates that certain mobile health apps can assist in cutting back on substance use, such as alcohol, at least in controlled environments. However, the real-world effectiveness of these apps is often compromised. App stores frequently promote revenue-generating apps over scientifically validated ones, resulting in a challenge for consumers to find evidence-based solutions.
Systematic reviews of substance use reduction apps consistently reveal a concerning trend: most apps fail to employ proven, evidence-based approaches. Instead, they make bold claims about their effectiveness and employ scientific-sounding language to enhance credibility, even if the underlying research is lacking.
Distinguishing Evidence-Based Apps:
Consumers can take proactive steps to identify evidence-based apps. Key indicators include:
- Scientific Research Citations: Apps that cite peer-reviewed studies or other scientific research are more likely to be reliable.
- Expert Collaboration: Development by experts in the field, often in collaboration with universities or licensed professionals, adds credibility.
- Independent Evaluation: Apps with published evaluations in scientific journals have undergone rigorous scrutiny.
- Data Standards and Compliance: Clear data storage practices and compliance with regulations like HIPAA ensure app integrity.
- Authenticity: Apps that avoid exaggerated promises and guarantee results are less likely to be misleading.
Regulatory Landscape and Concerns:
The current regulatory landscape is characterized by a significant lack of enforcement. The absence of substantiated health claims in mobile applications leaves a vast population vulnerable to misinformation, which can impede treatment and recovery for individuals with substance use disorders.
Generative AI's Role and Risks:
The integration of generative AI into health mobile apps has led to an influx of unregulated and untested products. While general-purpose models like ChatGPT suggest increased access to accurate health information, there are critical safety concerns. These models may provide inaccurate data, fail to respond appropriately in crisis situations, or normalize unsafe behaviors.
Protecting Consumers from Unregulated Apps:
Consumers are advised to exercise caution when encountering apps that use vague phrases like 'clinically proven' without providing specific details or references. Apps employing overly simplistic or implausible methods should also raise suspicion.
Strengthening Oversight of Generative AI:
One promising approach to regulating the health app marketplace is mandating Food and Drug Administration (FDA) approval. This would require apps to undergo randomized clinical trials and meet specific standards before becoming publicly available.
In the absence of stringent regulations, clear labeling is essential. Informed consumers need to distinguish between evidence-based and non-evidence-based apps. Effective safeguards, including fines, suspensions, and the removal of non-compliant products from app stores, are crucial to ensuring the accuracy, safety, and ethical use of mobile health apps.
Source and Further Reading:
For more insights, refer to the commentary titled 'The Need for Oversight Over Apps for Substance Use Reduction' published in the Journal of the American Medical Association (JAMA). The article, authored by Russell et al. (2025), is available at the following link: https://jamanetwork.com/journals/jama/article-abstract/2840575
