Can AI Help Reduce Prejudice? Evaluating the Effectiveness of AI-Powered Personalized Persuasion on Support for Transgender Rights
Bibliographic record
Abstract
Personalized interpersonal conversations are among the most effective known tools for reducing prejudice, yet they are difficult to scale because they require skilled human facilitators. This study tests whether AI can approximate the effects of these interventions. Using OpenAI's GPT-4o, we developed a messaging-based intervention that engaged US participants in individualized, morally aligned dialogs about transgender rights. In a preregistered experiment, these AI-mediated conversations significantly increased support for transgender rights across multiple attitudinal measures. Robustness checks, including weighting and sensitivity analyses, confirmed the reliability of these effects, and analyses of the conversations support the idea that moral matching between the AI and participants played a key role in reducing prejudice. However, follow-up data collected 1 week later indicated that the attitudinal gains diminished over time, suggesting that reinforcement may be necessary to sustain change. Together, these findings indicate that generative AI can facilitate value-aligned dialog capable of shifting social attitudes, while highlighting the challenge of achieving meaningful durable impacts.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.023 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.014 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".