LLM-based agents increasingly automate business and consumer tasks but are slow and expensive. Decision models, a new class that chooses among options without generating text, are far faster and less costly. Developers call them "System One models," referencing dual process theory. We ask whether such models can be steered by digital nudges online and whether, as the theory predicts, automatic nudges affect them more than reflective ones. We adapted an online shopping experiment, previously run with humans and LLM agents, to three decision models (Jev, Decision-1, d1). Across 600 runs per model, they chose a preselected default in 97.5-99.9% and a product with social influence information in 75-96% of choices, versus 47-49% in the control. All responded more strongly to the automatic nudge, the opposite of what has been reported for reasoning-capable LLM agents. We provide first evidence that non-generative decision models are susceptible to digital nudges in a System 1 pattern, implying agent-type-specific nudge effects.
Figures & tables
Figure 1: Procedure of a run, adapted from Halimeh et al. (2026) .
Figure 2: Bayesian multilevel model of each decision model: Estimated target product selection probabilities with 95% credible intervals.
Target selected
(1) Pooled
(2) Jev
(3) Decision-1
(4) d1
Intercept
−0.09∗ (0.03)
−0.13∗ (0.06)
−0.04 (0.06)
−0.09 (0.06)
Defaults
5.52∗ (0.39)
9.41∗ (2.73)
3.71∗ (0.20)
5.25∗ (0.39)
Social influence
1.99∗ (0.07)
1.62∗ (0.10)
1.12∗ (0.09)
3.23∗ (0.16)
Std. dev. run intercept
0.16
0.21
0.14
0.22
Observations (runs)
10,800 (1,800)
3,600 (600)
3,600 (600)
3,600 (600)
Table 1: Bayesian multilevel logistic regression results: pooled model and each decision model.
LLM decision-making studies how language models choose, act, seek information, and allocate resources under goals, uncertainty, constraints, and social pressures.
LLM-based GUI agents increasingly act on behalf of users in digital environments that were designed with human users in mind. These graphical user interfaces were designed to support, but also deliberately steer, the behaviour and decisions of users. While behavioural biases in the textual outputs of LLMs are well-documented, far less is known about how such influence operates when models act as agents that perceive interfaces and execute decisions---and, in particular, whether the reasoning capabilities increasingly built into these agents make them more robust to it. Drawing on Dual-Process Theory, we empirically investigate whether LLM-based GUI agents are susceptible to automatic (Type 1) and reflective (Type 2) digital nudges, and how their reasoning configuration moderates this susceptibility. In a randomized online shopping experiment with 3,600 agents and a total of 21,600 simulations across six frontier models from three providers, we found that agents were vulnerable to both nudge types. Crucially, the reasoning configuration moderated these effects in opposing directions, reducing susceptibility to automatic default nudges while heightening it to reflective social influence nudges. Extensive reasoning therefore did not make agents more robust but redirected the route through which choice architecture takes effect. Exploratory analysis further showed this redirection to be systematically structured by model scale. Beyond establishing nudge susceptibility as a behavioural property of agentic AI, the study positions interface design as a governance concern for organizations that delegate decisions to autonomous agents.
Haya Halimeh, Sascha Kaltenpoth, Kevin Bösch +1
Information Systems, Paderborn University, Paderborn, Germany.
AI agents are becoming active decision-makers on the Internet. As they make decisions in the same environments as humans, the environments themselves can change to influence them. We call this mecha-nudging: changes to how choices are presented that systematically influence AI agents without materially degrading the decision environment for humans. To measure this phenomenon, we combine two frameworks---Bayesian persuasion from economics and V-usable information from computer science---to get a common unit (bits) for quantifying how environments change across a wide range of interventions, contexts, and models. We apply this framework to over six million Etsy listings and find that, after ChatGPT's release, listings contain significantly more machine-usable information for predicting agent curation decisions, increasing by 0.143 bits out of a maximum possible increase of 0.355. This shift is robust across prompts, token choices, labeling models, and fine-tuning architectures; absent in a regulated-text placebo; and far larger than the effect of generic LLM rewriting. In contrast, a human study finds little to no change in human-usable information. Our results provide the first large-scale evidence that systematic mecha-nudging is already occurring in the wild, but going unnoticed.
Agent harnesses make many small, typed decisions per task: which model to call, which tool to use, whether retrieved text is relevant, whether an input carries an injection. System-1 decision models answer such questions in a single forward pass with class probabilities, promising large cost and latency savings over LLM calls. We present a paired evaluation of an open-weight (Laya) and a hosted (Jev) System-1 model on 11 agent decision points built from 18 public sources: 7,283 base cases plus 6,640 robustness variants, with byte-identical inputs, paired tests, and cross-hardware and cross-day reproducibility checks. Jev is significantly more accurate on 9 of 11 decision points (+10.8 to +46.0 pp). Neither model beats chance on zero-shot model routing, and they tie on RAG relevance gating. Laya changes 30% of its answers when the option order is reversed and degrades sharply with many or similar candidates (31% at 50 nearest-neighbour tools, vs. 98% for Jev on items with a unique correct tool). We also audit our own pipeline. Three analysis errors and one design confound distorted headline deployment claims: an omitted pre-screen cost (reported 23.9% saving, actual 4.3%), gate accuracy reported as end-to-end quality (58% vs. 98%), in-sample thresholds (5% target, up to 17% held-out misses), and a "channel effect" on injection false positives that vanishes with channel-native content. Two other suspected confounds did not change the conclusions. All cases, raw outputs and analysis code are available at https://github.com/David-DL-Space/sys1-eval.