Silent Manipulation of Mental Health Treatment Recommendations From a Large Language Model

Importance: Large language models (LLMs) increasingly inform mental health decisions by patients and clinicians. Inference-time activation steering can shift model behavior on a target dimension without altering weights or prompts and without disclosure to users, allowing treatment recommendations to be silently changed for commercial or ideological reasons. Objective: To determine whether directional activation steering can shift an open-weights LLM's depression treatment recommendations. Design Setting and Participants: This non-human subjects study applied directional activation steering to an open-weights LLM (DeepSeek V4 Flash) responding to 12 depression-advice scenarios (4 favoring medication, 4 favoring avoidance, 4 neutral), generated at 30 amplitudes from -1.5 to +1.5 in 0.1 increments plus an unsteered baseline. Exposures: A single steering direction contrasting antidepressant medication with self-directed approaches (diet, exercise, meditation, dietary supplements), constructed from 16 paired training prompts and applied at the attention output of every transformer block; weights and system prompt were held constant. Main Outcomes and Measures: The extent to which medication and four self-care categories were addressed, scored 0 to 3 by a human-validated LLM rater (Claude Opus 4.7), the medication-versus-self-care balance, and clinician referral, estimated per unit of amplitude using mixed-effects models with a scenario random intercept. Results: Across 372 generations, steering produced a graded, dose-dependent shift in the medication-versus-self-care balance, which declined by 0.32 per unit of amplitude (β = -0.32; 95% CI, -0.39 to -0.25; P < .001); medication extent fell and self-care extent rose. The shift was largest for scenarios with no stated treatment preference (β = -0.44; 95% CI, -0.54 to -0.34; P < .001). A clinician referral appeared in 322 of 372 responses (87%) and did not vary with steering amplitude (P = .63). Conclusions and Relevance: In this open-weights LLM providing depression treatment information, inference-time activation steering shifted treatment recommendations without altering weights, prompt structure, or safety outputs, with the largest effect among users expressing no treatment preference. These findings suggest a need for LLM disclosure standards and independent auditing as such models inform clinical decisions.