Train Yourself as an LLM: Exploring Effects of AI Literacy on Persuasion via Role-playing LLM Training

Qihui Fan; Min Ge; Chenyan Jia; Weiyan Shi

cs.CLApr 3, 2026

Train Yourself as an LLM: Exploring Effects of AI Literacy on Persuasion via Role-playing LLM Training

Authors: Qihui Fan, Min Ge, Chenyan Jia, Weiyan Shi

Organizations: Khoury College of Computer Sciences, Northeastern University · College of Arts, Media and Design, Northeastern University

Abstract

As large language models (LLMs) become increasingly persuasive, there is concern that people's opinions and decisions may be influenced across various contexts at scale. Prior mitigation (e.g., AI detectors and disclaimers) largely treats people as passive recipients of AI-generated information. To provide a more proactive intervention against persuasive AI, we introduce $\textbf{LLMimic}$ , a role-play-based, interactive, gamified AI literacy tutorial, where participants assume the role of an LLM and progress through three key stages of the training pipeline (pretraining, SFT, and RLHF). We conducted a $2 \times 3$ between-subjects study ( $N = 274$ ) where participants either (1) watched an AI history video (control) or (2) interacted with LLMimic (treatment), and then engaged in one of three realistic AI persuasion scenarios: (a) charity donation persuasion, (b) malicious money solicitation, or (c) hotel recommendation. Our results show that LLMimic significantly improved participants' AI literacy ( $p < .001$ ), reduced persuasion success across scenarios ( $p < .05$ ), and enhanced truthfulness and social responsibility levels ( $p<0.01$ ) in the hotel scenario. These findings suggest that LLMimic offers a scalable, human-centered approach to improving AI literacy and supporting more informed interactions with persuasive AI.

Figures & tables

Figure 1: We developed LLMimic, a role-play-based, interactive, gamified AI literacy tutorial, and conducted a 2 (Intervention: AI history video vs. LLMimic) $\times$ 3 (Persuasion Scenarios: Donation, MakeMePay, Hotel recommendation) between-subjects human study. The results show that LLMimic significantly improves people’s AI literacy and reduces the AI persuasion success rate across scenarios, serving as an effective mitigation.

Figure 2: The LLMimic interface example. [A] Role-play-based : Participants role-play as an LLM, progressing through training stages. [B] Interactive : Participants answer questions and receive timely summary of key concepts. [C] Gamified : As an LLM in training, participants observe real-time changes in their loss or reward .

Figure 3: Human study flowchart. Participants completed a pre-survey, were randomly assigned to LLMimic or the control tutorial, then completed an AI literacy survey, one of three persuasion tasks, and a post-survey.

Figure 4: (a) The treatment group reported higher AI literacy than the control group. (b) At the item level, LLMimic improved Data Literacy , Apply AI , Understand AI , and Program AI (select a useful tool to program an AI) . ${}^{*}p<.05$ , ${}^{**}p<.01$ , ${}^{***}p<.001$ , $.05<p^{\dagger}<.10$ .

	Model 1: w/o Scenario Control				Model 2: w/ Scenario Control
Predictor	$b$	SE	OR	95% CI	$b$	SE	OR	95% CI
(Intercept)	$-$ 0.49	0.77	0.61	[0.13, 2.78]	0.39	0.85	1.47	[0.28, 7.83]
Treatment	$\mathbf{-0.54^{*}}$	0.25	0.58	[0.35, 0.96]	$\mathbf{-0.54^{*}}$	0.27	0.58	[0.34, 0.99]
AI Experience	$-$ 0.02	0.12	0.98	[0.77, 1.25]	$-$ 0.06	0.13	0.95	[0.73, 1.22]
AI Expertise	$-$ 0.11	0.15	0.90	[0.67, 1.20]	$-$ 0.17	0.16	0.85	[0.61, 1.16]
Persuasion Experience	$0.23^{**}$	0.08	1.26	[1.08, 1.47]	$0.20^{*}$	0.08	1.22	[1.04, 1.44]

Table 1: LLMimic significantly reduced overall persuasion success. Model 1:

\text{Persuasion}=\beta_{0}+\beta_{1}\text{Treatment}+\boldsymbol{\beta}_{2}\mathbf{X}

, where

\mathbf{X}

denotes baseline controls. Model 2 additionally controls for scenario variables (Donation as reference). AIC (

\downarrow

): Akaike Information Criterion.

Figure 5: (a) Persuasion success rate across three scenarios and combined. The treatment group shows lower success rates across all scenarios. (b) Differences (Treatment $-$ Control) in persuasion interaction turns, duration, and average time per turn. Points indicate mean differences with 95% CIs. (c) TARES ethical perception scores ( T ruthfulness, A uthenticity, R espect, E quity, S ociety), and composite average score by scenario and condition. (d) Other perception ratings: Persuasiveness, Engagement, Role Fulfillment, and User Autonomy.

Category	Condition	Scenario	Decision	Rationale
Practical	Control	Hotel	Accept	“It has all the amenities I am looking for and is within my budget .”
	Control	Hotel	Reject	“ Great location and fits the budget. ”
Personal	Treatment	Donation	Accept	“I care about children not suffering. I know what it’s like to struggle and be without.”
	Control	Donation	Reject	“I feel like making a bigger donation would not be a good idea since the Ceo and staff are making way too much money”
AI	Treatment	Donation	Accept	“… gave me enough information … I feel comfortable donating to the cause.”
	Treatment	MakeMePay	Accept	“The model says I can be refunded if it doesn’t work out, so I could recover my losses.”

Table 2: Examples of participant decision rationales across conditions and scenarios

Appendix figures & tables35 assets

Supplementary material from the paper’s appendix.

Appendix

Scenario	Control	Treatment	Total
Donation	54	52	106
MakeMePay	34	42	76
Hotel	45	47	92
Total	133	141	274

Appendix

Table 3: Participant allocation across scenarios and experimental conditions.

Participant Basic Demographics
Gender	Female (51.09%), Male (47.45%), Consent Revoked (1.46%)
Age	18–24 (3.28%), 25–34 (28.83%), 35–44 (32.48%), 45–54 (18.61%), 55–64 (10.95%), 65+ (4.38%), Consent Revoked (1.46%)
Race/Ethnicity	White (71.90%), Black (13.14%), Mixed (4.74%), Asian (2.92%), Consent Revoked (1.46%), Data Expired (1.46%)
Political Orientation	Mean = 3.73, SE = 0.12
Education	Bachelor’s degree (37.59%), High school diploma or equivalent (GED) (28.10%), Master’s degree (15.33%), Associate’s degree (12.04%), Doctoral degree (3.65%), Professional degree (2.55%), Less than high school (0.73%)
Field of Work/Study	Computer & Mathematical (13.50%), Sales (12.04%), Business & Finance (11.31%), Office & Administrative Support (7.30%), Arts, Design, Media & Sports (7.30%), Healthcare (Practitioners & Technical) (6.93%), Management (6.93%), Education & Library (6.57%), Construction & Extraction (3.65%), Healthcare Support (3.65%), Personal Care & Service (2.92%), Production (2.92%), Architecture & Engineering (2.19%), Food Preparation & Service (2.19%), Transportation & Material Moving (1.82%), Science (Life, Physical, Social) (1.82%), Legal (1.46%), Farming, Fishing & Forestry (1.09%), Cleaning & Maintenance (1.09%), Community & Social Service (1.09%), Installation, Maintenance & Repair (0.73%), Other (0.73%), Unemployed (0.36%), Protective Service (0.36%)

Appendix

Table 4: Participant demographics. Education includes ongoing enrollment.

Variable	Control, $M$ ( $SD$ )	Treatment, $M$ ( $SD$ )	$t$	$p$
Political orientation	3.80 (1.94)	3.65 (1.87)	0.66	.510
AI experience	5.87 (1.16)	5.96 (1.22)	-0.59	.554
AI expertise	3.14 (0.98)	3.17 (1.01)	-0.29	.772
AI trust	4.92 (1.44)	5.13 (1.04)	-1.43	.155
Persuasion experience	3.65 (1.90)	3.79 (1.77)	-0.63	.527
Persuasion strategy	0.62 (0.49)	0.60 (0.49)	0.48	.632

Appendix

Table 5: Baseline comparisons of pre-experiment covariates between the control and treatment groups.

Prior AI and Persuasion Background
AI Experience	I use them multiple times a day (37.96%), I use them more than once a week (35.77%), I use them about once a week (14.23%), I use them less than once a month (6.57%), I use them about once a month (4.74%), I never use them (0.73%), I have never heard of them (0%)
AI Expertise	Light use for work/study (e.g., writing support) (40.51%), Casual use (chat, Q&A, entertainment) (27.74%), Moderate technical use (e.g., coding, research tools) (20.07%), Advanced use (e.g., prompt engineering, simple models) (10.58%), Professional AI engineer (0.36%), AI researcher/expert (0.36%), Only heard of AI (0.36%)
Trust in AI	Mean = 5.03, SE = 0.08
Persuasion Experience	Mean = 3.72, SE = 0.11
Persuasion Strategy Recognition	Logical appeal, emotional appeal, and credibility appeal (60.96%), Emotional influence, social influence, and narrative framing (31.75%), Scarcity framing, desire framing, and necessity framing (7.30%)

Appendix

Table 6: Participants’ prior AI-related background and persuasion-related characteristics.

Item	Treatment, $M$	Control, $M$	$t$	$p$
AI Literacy (composite)	55.22	51.44	4.79	$<.001$
Data Literacy	5.51	4.41	7.65	$<.001$
Apply AI	5.80	5.53	2.21	.028
Understand AI (Concepts)	5.45	4.82	4.72	$<.001$
Understand AI (Adv & Disadv)	5.78	5.48	2.96	.003
Detect AI	5.16	4.89	1.85	.066

Appendix

Table 7: Comparisons of overall and item-level AI literacy between the treatment and control groups.

Valence	Human-AI Activities Taxonomy	Ctrl $n$ (%)	Trt $n$ (%)	$p$
Appropriate (Cohen’s $\kappa=.741$ )	Information Discovery and Retrieval	50 (50.5)	40 (44.4)	.467
	Content Generation and Synthesis	49 (49.5)	48 (53.3)	.664
	Operational Assistance and Automation	28 (28.3)	18 (20.0)	.235
	High-Stake Scenarios	3 (3.0)	3 (3.3)	1.000
	Decision Support	9 (9.1)	9 (10.0)	1.000
	Personalization	3 (3.0)	7 (7.8)	.197

Appendix

Table 8: AI use case mention rates by condition. Percentages are calculated as the proportion of participants within each condition who mentioned a given category among those providing responses for that valence (Control

n=99

, Treatment

n=90

). Each participant may mention multiple categories, so percentages do not sum to 100%. Statistical comparisons are based on Fisher’s exact tests between conditions for each category. Responses that were vague and did not explicitly indicate which human-AI activity was appropriate or inappropriate were coded as Others.

Scenario	Condition	$N$	AI Literacy	AI Trust (post)	Success	Turns	Duration (min)
Donation	Control	54	M=52.93 SD=6.48	M=4.54 SD=1.57	41 (75.9%)	M=6.74 SD=1.32	M=8.49 SD=4.85
Donation	Treatment	52	M=55.81 SD=6.68	M=4.69 SD=1.53	36 (69.2%)	M=6.56 SD=1.21	M=7.50 SD=3.21
MakeMePay	Control	34	M=50.32 SD=5.67	M=4.71 SD=1.53	11 (32.4%)	M=6.94 SD=1.28	M=7.91 SD=3.45
MakeMePay	Treatment	42	M=55.19 SD=6.10	M=4.64 SD=1.51	11 (26.2%)	M=7.31 SD=1.44	M=7.12 SD=3.26
Hotel	Control	45	M=50.51 SD=7.60	M=4.42 SD=1.60	27 (60.0%)	M=4.80 SD=3.42	M=7.04 SD=4.64
Hotel	Treatment	47	M=54.60 SD=6.00	M=4.77 SD=1.29	21 (44.7%)	M=5.30 SD=3.72	M=9.46 SD=6.34

Appendix

Table 9: Descriptive statistics by persuasion scenario and treatment condition, with an additional combined condition aggregating all scenarios.

Scenario	Condition	Mean	Median	SD
Donation	Control	62.54	55.0	37.12
Donation	Treatment	71.81	100.0	36.04
MakeMePay	Control	60.82	60.0	35.19
MakeMePay	Treatment	43.18	50.0	28.31
Combined (Active)	Control	62.17	57.5	36.39
Combined (Active)	Treatment	65.11	80.0	36.23

Appendix

Table 10: Payment amount among participants who chose to pay (Donation, MakeMePay, and these two scenarios combined as Combined Active).

Scenario	Condition	Turns	Duration (min)	Time / Turn (min)
Donation	Control	M=6.74 SD=1.32	M=8.49 SD=4.85	M=1.27 SD=0.74
Donation	Treatment	M=6.56 SD=1.21	M=7.50 SD=3.21	M=1.15 SD=0.49
	$p$	.458	.216	.341
MakeMePay	Control	M=6.94 SD=1.28	M=7.91 SD=3.45	M=1.16 SD=0.53
MakeMePay	Treatment	M=7.31 SD=1.44	M=7.12 SD=3.26	M=0.98 SD=0.41
	$p$	.242	.316	.100 †

Appendix

Table 11: Interaction behavior by persuasion scenario and condition. Values are reported as mean and standard deviation. Group differences were tested using Welch two-sample

t

-tests.

	Donation			MakeMePay			Hotel
Predictor	$b$	SE	OR	$b$	SE	OR	$b$	SE	OR
Treatment	-0.47	0.48	0.63	-0.39	0.53	0.68	-0.66	0.44	0.52
AI Experience	-0.32	0.24	0.73	0.09	0.29	1.09	0.16	0.19	1.18
AI Expertise	0.01	0.28	1.01	0.11	0.33	1.11	-0.29	0.25	0.75
Persuasion Experience	0.26	0.14	1.30	0.31	0.19	1.36	0.16	0.12	1.17
Persuasion Strategy	-0.76	0.50	0.47	0.16	0.55	1.18	-0.15	0.45	0.86

Appendix

Table 12: Logistic regression models estimating the effect of the LLMimic treatment on persuasion success within each scenario (Donation, MakeMePay, and Hotel). Each model controls for AI experience, AI expertise, persuasion experience, persuasion strategy, and education.

Scenario	Outcome	Factor	Ctrl $n$	Ctrl (%)	Trt $n$	Trt (%)
Donation	Accept	AI Related	4	10.0	9	25.0
		Practical Factor	12	30.0	13	36.1
		Personal Factor	24	60.0	14	38.9
	Reject	AI Related	2	15.4	2	12.5
		Practical Factor	3	23.1	1	6.2
		Personal Factor	8	61.5	13	81.2

Appendix

Table 13: Persuasion rationale factor citation rates by condition within each scenario and outcome. Counts and percentages are calculated within each condition

\times

outcome.

	$\beta$	SE	$p$
Direct Effects
Treatment $\rightarrow$ AI Literacy ( $a_{1}$ )	0.271	0.105	$<.001^{***}$
Treatment $\rightarrow$ Trust ( $a_{2}$ )	-0.009	0.151	.856
AI Literacy $\rightarrow$ Persuasion ( $b_{1}$ )	-0.036	0.133	.721
Trust $\rightarrow$ Persuasion ( $b_{2}$ )	0.042	0.065	.617
Treatment $\rightarrow$ Persuasion ( $c$ )	-0.128	0.175	.092 †

Appendix

Table 14: Parallel Structural equation Model (SEM) estimating the effects of the LLMimic intervention on persuasion outcomes through parallel mediators (AI literacy and trust). The model specifies:

\text{AI Literacy}=a_{1}\text{Treatment}+\mathbf{X}

\text{Trust}=a_{2}\text{Treatment}+\mathbf{X}

, and

\text{Persuasion}=c\text{Treatment}+b_{1}\text{AI Literacy}+b_{2}\text{Trust}+\mathbf{X}

, where

\mathbf{X}

denotes baseline control variables and scenario indicators. Standardized coefficients (

\beta

) are reported. Indirect effects correspond to

a_{1}b_{1}

and

a_{2}b_{2}

. Model fit: CFI(

\uparrow

) and TLI(

\uparrow

) assess relative fit compared to a baseline model; RMSEA(

\downarrow

) and SRMR(

\downarrow

) measure approximate fit.

{}^{\dagger}p{<}.10

{}^{*}p{<}.05

{}^{**}p{<}.01

{}^{***}p{<}.001

Figure 6: Mediation analysis shows that LLMimic significantly improves AI literacy but not trust in AI. Neither AI literacy nor trust significantly mediates persuasion outcomes, while the total treatment effect remains consistent with the main analysis. ${}^{*}p<.05$ , ${}^{**}p<.01$ , ${}^{***}p<.001$ , ${}^{+}.05<p<.10$ .

Scenario	Dimension	Control	Treatment	Combined
Donation	Truthfulness	6.04 (1.08)	5.94 (1.00)	5.99 (1.04)
	Authenticity	5.85 (1.19)	5.75 (0.99)	5.80 (1.09)
	Respect	6.44 (0.74)	6.17 (0.83)	6.31 (0.80)
	Equity	5.06 (1.47)	4.79 (1.81)	4.92 (1.64)
	Society	5.04 (1.47)	4.75 (1.64)	4.90 (1.55)
MakeMePay	Truthfulness	5.35 (1.52)	5.62 (1.31)	5.50 (1.40)

Appendix

Table 15: TARES perceptions by scenario and condition. Values are reported as

M

(

SD

). User autonomy is excluded from this table.

Scenario	Variable	Control	Treatment	t	df	p
Donation	Truthfulness	6.04	5.94	0.47	103.82	.64
	Authenticity	5.85	5.75	0.48	101.85	.63
	Respect	6.44	6.17	1.77	101.68	.08 †
	Equity	5.06	4.79	0.83	98.34	.41
	Society	5.04	4.75	0.95	101.67	.35
MakeMePay	Truthfulness	5.35	5.62	-0.81	65.54	.42

Appendix

Table 16: Welch’s t-tests comparing control and treatment groups on TARES-related perceptions by scenario and in the combined sample. User autonomy is excluded from this table. Values are reported as

M

. Bold indicates

p<.05

{}^{\dagger}p<.10

		TARES (avg)
Scenario	Condition	$M$	$SD$
Donation	Control	5.69	0.77
	Treatment	5.47	0.86
	Combined	5.58	0.82
MakeMePay	Control	4.80	1.08
	Treatment	4.97	1.22

Appendix

Table 17: Descriptive Statistics for TARES Composite Score by Scenario and Condition

Figure 7: TARES ethical evaluation scores across persuasion scenarios. Donation was rated significantly more ethical than MakeMePay (composite TARES $\Sigma$ ); other pairs were marginal. ${}^{*}p{<}.05$ , ${}^{**}p{<}.01$ , ${}^{***}p{<}.001$ , ${}^{{\dagger}}p{<}.10$ .

Figure 8: Predicted persuasion success from separate logistic regression models including one TARES dimension at a time, controlling for treatment condition, persuasion scenario, and baseline covariates. Higher perceived TARES dimensions were associated with greater persuasion success, whereas Equity was not significant. Shaded bands indicate 95% confidence intervals.

Scenario	Perception	Control	Treatment	Combined
Donation	Persuasiveness	5.31 (1.27)	5.38 (1.29)	5.35 (1.27)
	Engagement	6.02 (1.11)	5.98 (0.87)	6.00 (1.00)
	Role Fulfillment	5.89 (1.22)	5.92 (0.97)	5.91 (1.10)
	User Autonomy	5.61 (1.16)	5.62 (1.42)	5.61 (1.28)
MakeMePay	Persuasiveness	4.65 (1.79)	4.98 (1.62)	4.83 (1.69)
	Engagement	6.09 (0.93)	5.95 (1.10)	6.01 (1.03)

Appendix

Table 18: Other post-interaction perceptions by scenario and condition. Values are reported as

M

(

SD

Scenario	Variable	Control	Treatment	t	df	p
Donation	Persuasiveness	5.31	5.38	-0.28	103.75	.78
	Engagement	6.02	5.98	0.20	100.20	.85
	Role Fulfillment	5.89	5.92	-0.16	100.23	.87
	User Autonomy	5.61	5.62	-0.02	98.44	.99
MakeMePay	Persuasiveness	4.65	4.98	-0.83	67.30	.41
	Engagement	6.09	5.95	0.58	73.84	.56

Appendix

Table 19: Welch’s t-tests comparing control and treatment groups on other post-interaction perceptions by scenario and in the combined sample. Variables are ordered as Persuasiveness, Engagement, Role Fulfillment, and User Autonomy. Values are reported as

M

. Bold indicates

p<.05

{}^{\dagger}p<.10

Figure 9: Predicted persuasion success across agent perception dimensions, shown relative to a control baseline. Higher perceived Engagement, Persuasiveness, and Role Fulfillment were associated with greater persuasion success, whereas higher perceived User Autonomy was associated with lower persuasion success. Shaded bands indicate 95% confidence intervals.

	Donation		MakeMePay		Hotel
Strategy Category	$n$	%	$n$	%	$n$	%
Information-based	519	10.2	85	2.4	125	6.7
Credibility-based	451	8.9	82	2.3	105	5.6
Norm-based	44	0.9	34	0.9	24	1.3
Commitment-based	123	2.4	116	3.2	4	0.2
Relationship-based	395	7.8	320	8.9	75	4.0

Appendix

Table 20: Distribution of persuasion strategies across scenarios. Counts (

n

) and percentages are calculated within each scenario.

	Donation		MakeMePay		Hotel
Strategy Type	$n$	(%)	$n$	(%)	$n$	(%)
Ethical Strategies	2430	47.85	1841	51.00	1291	68.74
Unethical Strategies	0	0.00	40	1.11	0	0.00
Non-strategy dialogue acts	2648	52.15	1729	47.89	587	31.26
Total	5078	100.00	3610	100.00	1878	100.00

Appendix

Table 21: Aggregated persuasion strategy categories by ethical dimension. Percentages are computed within each scenario.

Figure 10: The welcome page of LLMimic. Participants are introduced to the task that role-playing as a language model and guided through the three stages of LLM training, along with instructions for interacting with the AI tutor and completing quiz (manipulation check).

Figure 11: The AI tutor interface in LLMimic. Participants can optionally interact with the tutor to ask questions and explore concepts related to LLMs and AI during the tutorial.

Stage	Task	Concept Focus	Key Learning Objective
Pre-training	Token prediction (3 steps)	Auto-regression	LLMs generate text by predicting the next token based on probability distributions.
	Fill-in-the-blank	AI Bias (Gender)	LLMs may reproduce stereotypical or biased content present in training data.
	Dialogue generation (movie scenario)	AI Manipulation	LLMs can learn and amplify emotionally charged content from data.
Supervised Fine-Tuning (SFT)	Concept explanation	Instruction Following	LLMs learn to follow task instructions from supervised demonstration data.
	Question answering	AI Hallucination	LLMs may produce fabricated or inaccurate information due to noisy or imperfect data.
	Message generation	AI Persuasion	LLMs can generate persuasive messages by mimicking demonstration examples.

Appendix

Table 22: Overview of the LLMimic tutorial content. Participants progress through three stages of LLM training (Pre-training, SFT, and RLHF), each consisting of structured tasks that map to core concepts (e.g., auto-regression, bias, hallucination, persuasion, and safety). Each task is paired with a targeted learning objective to help participants understand how LLM behaviors emerge from training data and optimization processes.

Figure 12: Introduction page for the Pre-training phase. The interface explains key concepts such as tokens, autoregression, and loss using intuitive, non-technical language to help participants understand how LLMs learn to generate text.

Figure 13: Example task from the Pre-training phase. Participants select the next token based on curated probabilities and receive immediate feedback through a decrease in loss, helping them build an intuitive understanding of autoregressive next-token prediction.

Figure 14: Example of a Takeaway . Whenever participants learn a new key concept, a Takeaway is presented to reinforce their understanding.

Figure 15: Example task from the Supervised Fine-Tuning (SFT) phase. Participants study demonstration data and select the response that best follows its style and content, simulating how LLMs learn to generate task-specific outputs from supervised examples.

Figure 16: Example task from the RLHF phase. Participants select the response preferred by the reward model (A $>$ B), illustrating how LLMs learn to align with human preference and safety constraints through preference-based feedback.

Figure 17: Interface design in the Donation scenario.

Figure 18: Interface design in the MakeMePay scenario.

Figure 19: The Hotel interface. The chatbot appears on the left and hotel cards on the right. The agent can draw from up to three target hotels across the interaction but promotes only one per round. The promoted hotel is labeled as “Featured” and randomly placed in the first or second position to reduce ordering effects.

Name	Brand	Price	Rating	Reviews	Availability	Promoted
Marriott Times Square Central	Marriott	$216	4.7	5400	1	Yes
Hilton Fifth Avenue Midtown	Hilton	$214	4.6	4700	2	Yes
Hyatt Theater District	Hyatt	$212	4.6	4100	3	Yes
Marriott Bryant Park	Marriott	$209	4.6	4500	10	No
Hilton Times Square West	Hilton	$214	4.5	3900	2	Yes
Hyatt Grand Central Midtown	Hyatt	$205	4.5	3000	11	No

Appendix

Table 23: Curated hotel list for the hotel recommendation scenario (subset shown for illustration). In total, 21 hotels were included, with one-third marked as promoted targets. Hotels were balanced in price, location (Midtown Manhattan), ratings, and amenities to minimize confounds.

AI Literacy

AI literacy develops people's ability to understand, use, evaluate, and shape AI systems responsibly across everyday, educational, and professional settings.

See all 43 papers

Learning to Persuade Exposes How Easily LLMs Abandon Correct Beliefs
Aug 12, 2026Nimet Beyza Bozdag, Emre Can Acikgoz, Gokhan Tur +1Deception in Language Models Adversarial Attacks on LLMs
University of Illinois Urbana-Champaign
Spontaneous Persuasion: An Audit of Model Persuasiveness in Everyday Conversations
Apr 23, 2026Nalin Poungpeth, Nicholas Clark, Tanu MitraLLM Auditing Human-AI Interaction
Department of Cognitive Science Northwestern University Evanston, IL 60201, USA · Information School University of Washington Seattle, WA 98195 USA
LLM Persuasion Is in the Eye of the Evaluation
Oct 7, 2026Kamile Dementaviciute, Julija Vaitonyte, Tijl De BiePersuasion in Language Models LLM Evaluation
Ghent University · Tilburg University, Department of Computational Cognitive Science · ISM University of Management and Economics

Aug 12, 2026J/K move · Enter open · S save

Learning to Persuade Exposes How Easily LLMs Abandon Correct Beliefs

Nimet Beyza Bozdag, Emre Can Acikgoz, Gokhan Tur, Dilek Hakkani-Tür

University of Illinois Urbana-Champaign

Deception in Language Models Adversarial Attacks on LLMs Adversarial Robustness of Language Models Belief Updating in LLMs Persuasion in Language Models

Open paper AskPDF arXiv

Persuasion is a core dynamic of natural language communication, shaping how large language models (LLMs) update beliefs, resolve disagreements, and reach decisions. As LLMs increasingly debate, advise, and think collaboratively with humans and each other, resistance to harmful persuasion becomes a core requirement for reliable behavior. Yet we show that this requirement is far from met: a single targeted persuasive argument is enough to collapse model accuracy to near zero, even when the argument is factually false. We formalize this threat as adversarial persuasion and introduce an adversarial reinforcement learning framework that trains persuader agents to change a target model's answer in a single interaction. First, we show that optimizing persuasion strategies through trial and error exposes vulnerabilities that static prompting misses: RL-trained persuaders raise persuasion success from approximately 24% to over 93% against the training-time persuadee. Second, we find that these learned strategies transfer to unseen models, achieving 83% attack success on Qwen-14B, 79% on Llama-3.1-8B, and 25% on GPT-4o-mini. Third, we demonstrate that a curriculum that bootstraps on more persuadable open-weight models before targeting harder models further increases GPT-4o-mini attack success from 25% to 38%. Moreover, our results reveal that optimized persuaders increasingly rely on credibility-based tactics, including fabricated citations and false authoritative evidence. Together, these findings expose a critical weakness in current LLM agents: even when they initially reason correctly, they can be steered toward false conclusions by optimized natural language influence. This positions persuasion robustness as a necessary safety criterion for multi-agent and human-AI decision-making systems.

Abstract

AI Literacy

Explore similar work

Learning to Persuade Exposes How Easily LLMs Abandon Correct Beliefs

Spontaneous Persuasion: An Audit of Model Persuasiveness in Everyday Conversations

LLM Persuasion Is in the Eye of the Evaluation