The current political and regulatory discourse frequently references the term “trustworthy artificial intelligence (AI).” In Europe, the attempts to ensure trustworthy AI started already with the High‐Level Expert Group Ethics Guidelines for Trustworthy AI and have now merged into the regulatory discourse on the EU AI Act. Around the globe, policymakers are actively pursuing initiatives—as the US Executive Order on Safe, Secure, and Trustworthy AI, or the Bletchley Declaration on AI showcase—based on the premise that the right regulatory strategy can shape trust in AI. To analyze the validity of this premise, we propose to consider the broader literature on trust in automation. On this basis, we constructed a framework to analyze 16 factors that impact trust in AI and automation more broadly. We analyze the interplay between these factors and disentangle them to determine the impact regulation can have on each. The article thus provides policymakers and legal scholars with a foundation to gauge different regulatory strategies, notably by differentiating between those strategies where regulation is more likely to also influence trust on AI (e.g., regulating the types of tasks that AI may fulfill) and those where its influence on trust is more limited (e.g., measures that increase awareness of complacency and automation biases). Our analysis underscores the critical role of nuanced regulation in shaping the human‐automation relationship and offers a targeted approach to policymakers to debate how to streamline regulatory efforts for future AI governance.

Background: Insufficient physical activity and unhealthy diets are contributing to the rise in noncommunicable diseases. Preventative mobile health (mHealth) interventions may help reverse this trend, but present bias might reduce their effectiveness. Future-self avatar interventions have resulted in behavior change in related fields, yet evidence of whether such interventions can change health behavior is lacking.Objective: We aimed to investigate the impact of a future-self avatar mHealth intervention on physical activity and food purchasing behavior and examine the feasibility of a novel automated nutrition tracking system. We also aimed to understand how this intervention impacts related attitudinal and motivational constructs.Methods: We conducted a 12-week parallel randomized controlled trial (RCT), followed by semistructured interviews. German-speaking smartphone users aged ≥18 years living in Switzerland and using at least one of the two leading Swiss grocery loyalty cards, were recruited for the trial. Data were collected from November 2020 to April 2021. The intervention group received the FutureMe intervention, a physical activity and food purchase tracking mobile phone app that uses a future-self avatar as the primary interface and provides participants with personalized food basket analysis and shopping tips. The control group received a conventional text- and graphic-based primary interface intervention. We pioneered a novel system to track nutrition by leveraging digital receipts from loyalty card data and analyzing food purchases in a fully automated way. Data were consolidated in 4-week intervals, and nonparametric tests were conducted to test for within- and between-group differences.Results: We recruited 167 participants, and 95 eligible participants were randomized into either the intervention (n=42) or control group (n=53). The median age was 44 years (IQR 19), and the gender ratio was balanced (female 52/95, 55%). Attrition was unexpectedly high with only 30 participants completing the intervention, negatively impacting the statistical power. The FutureMe intervention led to small statistically insignificant increases in physical activity (median +242 steps/day) and small insignificant improvements in the nutritional quality of food purchases (median −1.28 British Food Standards Agency Nutrient Profiling System Dietary Index points) at the end of the intervention. Intrinsic motivation significantly increased (P=.03) in the FutureMe group, but decreased in the control group. Outcome expectancy directionally increased in the FutureMe group, but decreased in the control group. Leveraging loyalty card data to track the nutritional quality of food purchases was found to be a feasible and accepted fully automated nutrition tracking system.Conclusions: Preventative future-self avatar mHealth interventions promise to encourage improvements in physical activity and food purchasing behavior in healthy population groups. A full-powered RCT is needed to confirm this preliminary evidence and to investigate how future-self avatars might be modified to reduce attrition, overcome present bias, and promote sustainable behavior change.Trial Registration: ClinicalTrials.gov NCT04505124; https://clinicaltrials.gov/ct2/show/NCT04505124

In order to induce the shift in consumer behavior necessary for the mitigation of diet-related diseases, front-of-package labels (FoPL) such as the Nutri-Score that support consumers in their efforts to identify nutritionally valuable products during grocery shopping have been found to be effective; however, they remain non-compulsory in most regions. Counter-intuitively, a similar stream of research on digital web-based FoPL does not yet exist, even though such digital labels hold several advantages over physical labels. Digital FoPL can provide scalable and personalized interventions, are easier to implement than physical labels, and are especially timely due to the recent increase in online grocery shopping. The goal of this study was to demonstrate the technical feasibility and intervention potential of novel, scalable, and passively triggered health behavior interventions distributed via easy-to-install web browser extensions designed to support healthy food choices via the inclusion of digital FoPL in online supermarkets. To that end, we developed a Chrome web browser extension for a real online supermarket and evaluated the effect of this digital food label intervention (i.e., display of the Nutri-Score next to visible products) on the nutritional quality of individuals’ weekly grocery shopping in a randomized controlled laboratory trial (N = 135). Compared to the control group, individuals exposed to the intervention chose products with a higher nutritional quality (e.g., 8% higher healthy trolley index (HETI), 3.3% less sugar, 7.5% less saturated fat). In particular, users with low food literacy seemed to benefit from the digital FoPL (e.g., 11% higher HETI, 10.5% less sugar, 5.5% less saturated fat). Furthermore, participants exposed to the food label advocated its introduction more strongly than the control group (p = 0.081). Consumers worldwide could easily install such applications to display digital food labels on their end devices, and would thus not have to wait for stakeholders in the food industry to eventually reach consensus on mandatory food label introduction.

For the Internet of Things to be adopted in people's homes and at their workplaces, it is important to provide mechanisms that support them when controlling and monitoring smart things in their surroundings. We present the concept of user interface beaming, where the capabilities of different personal wearable computers are combined to allow users to conveniently interact with smart things in their environment. Smartglasses are used to select a target smart thing by means of current object recognition technologies. Then, an appropriate user interface for the target is rendered on the user's smartwatch. This interface is continuously updated to reflect state changes of the target and can be used to interact with that smart thing using different interaction modalities.