A scoping review on the mental health harms of LLM-based chatbots - npj Digital Medicine - Nature

IntroductionHistoric masonry minarets are slender, tall structures that embody cultural identity while remaining vulnerable to progressive deterioration and sudden hazards1,2. Aging mortar, material heterogeneity, environmental exposure, nearby construction, and seismic activity can introduce cracks, detachments, and out-of-plane deformations that are difficult to detect early, especially when access is limited and interventions must be minimally invasive. Structural health monitoring (SHM) for such heritage assets therefore benefits from non-contact sensing and data-driven inference that can operate continuously, localize damage, and provide interpretable evidence to support conservation decisions. For tall, slender masonry forms, monitoring is additionally complicated by geometric discontinuities (e.g., balconies, openings, and transitions), spatially varying material properties, and complex soil–structure interaction, all of which can concentrate damage and alter dynamic response over time3,4,5.

“How are you doing?” “Are you sleeping well?” “Do you drink?” These are typical opening questions a mental-health professional might ask a patient. Soon, another question may be added to the list: “How do you use artificial intelligence (AI)?”

As more and more people turn to AI chatbots for quick, 24/7 medical advice, a new perspective paper co-authored by faculty at Binghamton University explores the risks this practice carries that are largely invisible to users – and healthcare systems. The paper, written by researchers at Binghamton University, Stanford University, Texas A&M University, and Indiana University, and published in Nature Health, examines what happens when a chatbot gives someone bad health advice. The authors argue that the harm is hard to catch because the conversation stays on the AI company's platform, the person has no place to report the problem, and outside researchers cannot measure the damage. “We’re getting medical advice from these chatbots, but nobody—literally no one—is looking at this. If there’s an error or an issue, there’s just no way for people to know,” said Kaicheng Yang, an assistant professor in the School of Computing at Binghamton University’s Thomas J. Watson College of Engineering and Applied Science and one of the paper’s authors. The researchers gave the example of a 60-year-old man who asked ChatGPT how to cut chloride from his diet and swapped table salt with sodium bromide.

Our security assessment of LiteLLM (v1.77.7), a widely used open-source LLM gateway, uncovered multiple high-impact vulnerabilities (tracked as CVE-2026-35029, CVE-2026-59823) which have been resolved as of LiteLLM version 1.83.9. The most significant issues stemmed from a missing authorization check on the /config/update endpoint, which allowed any authenticated user (including low-privileged accounts) to modify the proxy’s runtime configuration. This single flaw enabled arbitrary file read, administrative account takeover, and remote code execution via malicious Python modules loaded from cloud storage. A separate SSRF vulnerability in the “Clientside LLM Credentials” feature allowed attackers to redirect API calls and exfiltrate provider keys. These findings highlight that as AI infrastructure tools rapidly evolve, foundational security practices like proper access control and input validation remain essential.