Real-world adoption of closed-loop insulin delivery systems (CLIDS) in type 1 diabetes remains low, driven not by technical failure, but by diverse behavioral, psychosocial, and social barriers. We introduce ChatCLIDS, the first benchmark to rigorously evaluate LLM-driven persuasive dialogue for health behavior change. Our framework features a library of expert-validated virtual patients, each with clinically grounded, heterogeneous profiles and realistic adoption barriers, and simulates multi-turn interactions with nurse agents equipped with a diverse set of evidence-based persuasive strategies. ChatCLIDS uniquely supports longitudinal counseling and adversarial social influence scenarios, enabling robust, multi-dimensional evaluation. Our findings reveal that while larger and more reflective LLMs adapt strategies over time, all models struggle to overcome resistance, especially under realistic social pressure. These results highlight critical limitations of current LLMs for behavior change, and offer a high-fidelity, scalable testbed for advancing trustworthy persuasive AI in healthcare and beyond.

Background: Type 1 diabetes (T1D) has seen a rapid evolution in management technology and forms a useful case study for the future management of other chronic conditions. Further development of this management technology requires an exploration of its real-world use and the potential of additional data streams. To facilitate this, we contribute the BrisT1D Dataset to the growing number of public T1D management datasets. The dataset was developed from a longitudinal study of 24 young adults in the UK who used a smartwatch alongside their usual T1D management. Findings: The BrisT1D dataset features both device data from the T1D management systems and smartwatches used by participants, as well as transcripts of monthly interviews and focus groups conducted during the study. The device data is provided in a processed state, for usability and more rapid analysis, and in a raw state, for in-depth exploration of novel insights captured in the study. Conclusions: This dataset has a range of potential applications. The quantitative elements can support blood glucose prediction, hypoglycaemia prediction, and closed-loop algorithm development. The qualitative elements enable the exploration of user experiences and opinions, as well as broader mixed-methods research into the role of smartwatches in T1D management.

High quality real world datasets are essential for advancing data driven approaches in type 1 diabetes (T1D) management, including personalized therapy design, digital twin systems, and glucose prediction models. However, progress in this area has been limited by the scarcity of publicly available datasets that offer detailed and comprehensive patient data. To address this gap, we present AZT1D, a dataset containing data collected from 25 individuals with T1D on automated insulin delivery (AID) systems. AZT1D includes continuous glucose monitoring (CGM) data, insulin pump and insulin administration data, carbohydrate intake, and device mode (regular, sleep, and exercise) obtained over 6 to 8 weeks for each patient. Notably, the dataset provides granular details on bolus insulin delivery (i.e., total dose, bolus type, correction specific amounts) features that are rarely found in existing datasets. By offering rich, naturalistic data, AZT1D supports a wide range of artificial intelligence and machine learning applications aimed at improving clinical decision making and individualized care in T1D.

More than 150,000 adults and children with type 1 diabetes are now eligible to get an'artificial pancreas'. NHS regulators have today approved hybrid closed-loop system technology, which experts say is the'biggest breakthrough since insulin'. The high-tech device continuously tracks blood sugar levels through a sensor stuck to the body. Readings are fed straight back to a body-worn insulin pump, with an algorithm then calculating how much of the hormone needs to be released. An artificial pancreas to manage type 1 diabetes could soon be offered to NHS patients after a major trial produced'blisteringly brilliant' early results.

A fascinating session at the EASD 2022 conference on emerging technologies shed light on where we are with AID and telemedicine, and what leading researchers in diabetes believe is coming next in diabetes management. Healthcare is rapidly evolving, and now more than ever, robots and artificial intelligence have gone from science fiction to critical components of diabetes management. At the EASD 2022 conference in Stockholm, Sweden, researchers further explored this concept in a session titled, "A New Hope or Strange New Worlds: Submerging diabetes into emerging technologies." Dr. Moshe Phillip, head of the Institute of Endocrinology and Diabetes at Schneider Children's Medical Center of Israel, began by demonstrating how continuous glucose monitors (CGMs) represent a paradigm shift in diabetes technology. "CGM is the most important tool in the last 20 years," he said.

Despite the increasing adoption of insulin pumps and continuous glucose monitoring devices, most people with type 1 diabetes do not achieve their glycemic goals1. This could be related to a lack of expertise or inadequate time for clinicians to analyze complex sensor-augmented pump data. We tested whether frequent insulin dose adjustments guided by an automated artificial intelligence-based decision support system (AI-DSS) is as effective and safe as those guided by physicians in controlling glucose levels. ADVICE4U was a six-month, multicenter, multinational, parallel, randomized controlled, non-inferiority trial in 108 participants with type 1 diabetes, aged 10–21 years and using insulin pump therapy (ClinicalTrials.gov no. NCT03003806). Participants were randomized 1:1 to receive remote insulin dose adjustment every three weeks guided by either an AI-DSS, (AI-DSS arm, n = 54) or by physicians (physician arm, n = 54). The results for the primary efficacy measure—the percentage of time spent within the target glucose range (70–180 mg dl−1 (3.9–10.0 mmol l−1))—in the AI-DSS arm were statistically non-inferior to those in the physician arm (50.2 ± 11.1% versus 51.6 ± 11.3%, respectively, P < 1 × 10−7). The percentage of readings below 54 mg dl−1 (<3.0 mmol l−1) within the AI-DSS arm was statistically non-inferior to that in the physician arm (1.3 ± 1.4% versus 1.0 ± 0.9%, respectively, P < 0.0001). Three severe adverse events related to diabetes (two severe hypoglycemia, one diabetic ketoacidosis) were reported in the physician arm and none in the AI-DSS arm. In conclusion, use of an automated decision support tool for optimizing insulin pump settings was non-inferior to intensive insulin titration provided by physicians from specialized academic diabetes centers. The randomized-controlled trial ADVICE4U demonstrates non-inferiority of an automated AI-based decision support system compared with advice from expert physicians for optimal insulin dosing in youths with type 1 diabetes.

Most people with diabetes need at least two shots of insulin per day, but to to ease this burden, scientists are working on an implantable robot to administer the medication. A team of Italian researchers recently published a study in the journal Science Robotics that outlines a two-component system called PILLSID, which includes an implantable insulin pump that sits in the abdomen area and ingestible magnetic hormone capsules to refill it. When patients need to reload the pump, they swallow a capsule, which is then pulled through the digestive system by magnets inside the insulin device. The device, roughly the size of a flip phone, catches the capsule with a tractable needle, rotates it into a certain position and then extracts the hormone. The capsule continues to move naturally through the digestive track and eventually leaves the body.

As artificial intelligence algorithms and systems become more sophisticated and take on bigger responsibilities, it becomes more and more important to ensure that AI systems avoid dangerous, unwanted behavior. Recently a team of researchers from the University of Massachusetts Amherst and Stanford published a paper that demonstrates how specific AI behavior can be avoided, through the use of a technique that elicits precise mathematical instructions that can be used to tweak the behavior of an AI. According to TechXplore, the research was predicated on the assumption that unfair/unsafe behaviors can be defined with mathematical functions and variables. If this is true then it should be possible for researchers to train systems to avoid these specific behaviors. The research team aimed to develop a toolkit that could be employed by users of the AI to specify which behaviors they want the AI to avoid, and enable AI engineers to reliably train a system that will avoid unwanted actions when used in real-world scenarios.

Guaranteeing safe and fair machine behavior is still an issue today, says machine learning researcher and lead author Philip Thomas at the University of Massachusetts Amherst. "When someone applies a machine learning algorithm, it's hard to control its behavior," he points out. This risks undesirable outcomes from algorithms that direct everything from self-driving vehicles to insulin pumps to criminal sentencing, say he and co-authors. Writing in Science, Thomas and his colleagues Yuriy Brun, Andrew Barto and graduate student Stephen Giguere at UMass Amherst, Bruno Castro da Silva at the Federal University of Rio Grande del Sol, Brazil, and Emma Brunskill at Stanford University this week introduce a new framework for designing machine learning algorithms that make it easier for users of the algorithm to specify safety and fairness constraints. "We call algorithms created with our new framework'Seldonian' after Asimov's character Hari Seldon," Thomas explains.

Artificial intelligence, it seems, can figure out how to do just about anything. It can simulate the Universe, learn to solve a Rubik's Cube with just one hand, and even find ghosts hidden in our past. All these kinds of advancements are meant to be for our own good. In recent times, algorithmic systems that already affect people's lives have demonstrated alarming levels of bias in their operation, doing things like predicting criminality along racial lines and determining credit limits based on gender. Against this backdrop, how can scientists ensure that advanced thinking systems can be fair, or even safe?