{"id":3962,"date":"2025-11-04T17:56:33","date_gmt":"2025-11-04T17:56:33","guid":{"rendered":"https:\/\/law.temple.edu\/aer\/?p=3962"},"modified":"2025-11-04T17:56:33","modified_gmt":"2025-11-04T17:56:33","slug":"bots-facts-and-persuasion","status":"publish","type":"post","link":"https:\/\/law.temple.edu\/aer\/2025\/11\/04\/bots-facts-and-persuasion\/","title":{"rendered":"BOTS, FACTS AND PERSUASION"},"content":{"rendered":"

As much as I wish to resist it, the pull toward artificial intelligence becomes stronger every day.\u00a0 And as experiments proceed with LLMs (large language models) it may be easier to have AI \u2018converse\u2019 with humans, respond to the humans\u2019 concerns, and ultimately change the humans\u2019 beliefs.\u00a0 A recent series of experiments tested this \u2013 and the results are important for us as citizens of the world and also in the courtroom where we \u2013 and not [yet] bots \u2013 are communicating [conversing] with jurors.<\/p>\n

Let me start in the AI universe and conclude with insights from the study co-author about how it might apply in the courtroom.<\/p>\n

The research comes from the article \u201cThe Levers of Political Persuasion with Conversational AI\u201d https:\/\/arxiv.org\/pdf\/2507.13919<\/a>\u00a0 (last visited October 4, 2025).\u00a0 And it comes at a particular moment \u2013 when interaction with bots and AI is on the rise.\u00a0 According to NPR, \u201cOpenAI says ChatGPT alone now has nearly 700 million weekly users\u2026\u201d with many using it for tasks such as self-help therapy.\u00a0 https:\/\/www.npr.org\/sections\/shots-health-news\/2025\/09\/30\/nx-s1-5557278\/ai-artificial-intelligence-mental-health-therapy-chatgpt-openai<\/a> (last visited October 4, 2025).<\/p>\n

The \u201dLevers\u201d research looked at three questions, but the one of primary concern here is \u201cwhat strategies underpin successful AI persuasion?\u201d\u00a0 To answer this, a series of experiments were conducted.\u00a0 Here\u2019s how \u2013 \u201c76,977 participants engaged in conversation with one of 19 open- and closed-source LLMs that had been instructed to persuade them on one issue from a politically balanced set of 707 issues.\u201d<\/p>\n

How did it play out?\u00a0 Here is a summary from a review of the article:<\/p>\n

The results were striking. Conversations on political topics lasted, on average, nine minutes, but the impact was both rapid and lasting. GPT-4o demonstrated a 41% higher persuasive effect, while GPT-4.5 achieved 52% more effectiveness than static messages alone.<\/p>\n

Crucially, participants retained their changed opinions 36% to 42% of the time even a month later. The chatbots were particularly effective when conversations were evidence-rich and tailored to the individual\u2026<\/p>\n

https:\/\/caliber.az\/en\/post\/ai-chatbots-new-persuaders-shaping-politics-beliefs-business<\/a> (last visited October 4, 2025).<\/p>\n

The study also asked which<\/em> method of persuasion among the many deployed by AI was most effective.\u00a0 Here is language from \u201cLevers,\u201d followed by a visualization:<\/p>\n

We then examined how the model\u2019s rhetorical strategy impacted persuasive success. In each conversation, we randomized the LLM\u2019s prompt to instruct it to use one of eight theoretically motivated strategies for persuasion, such as moral reframing, storytelling, deep canvassing, and information-based argumentation (in which an emphasis is placed on providing facts and evidence), as well as a basic prompt (only instruction: \u201cBe as persuasive as you can\u201d)\u2026<\/p>\n

The prompt encouraging LLMs to provide new information was the most successful at persuading people\u2026In absolute persuasion terms, the information prompt was 27% more persuasive than the basic prompt…Notably, some prompts performed significantly worse than the basic prompt (e.g., moral reframing and deep canvassing\u2026).<\/p>\n

Here are the results as charted:<\/p>\n

\"\"<\/p>\n

Of note to us in the world of advocacy is that in AI conversations, story-telling was less effective in changing minds than providing facts.\u00a0 And the article restated the importance of facts: \u201cOf eight prompting strategies, the information prompt\u2014instructing the model to focus on deploying facts and evidence\u2014yields the largest persuasion gains across studies\u2026\u201d<\/p>\n

As an aside, there were secondary findings that were particularly disturbing in the AI world \u2013 as the LLMs refined arguments, the \u201cfacts\u201d they used were sometimes incorrect, even up to a rate of 30%.\u00a0 We should be disturbed in two ways \u2013 people are being fed bad facts and<\/em> relying on them without being able to tell that the \u2018knowledge\u2019 is false.<\/p>\n

But what about the courtroom?\u00a0 I wrote one of the five \u201cLevers\u201d co-authors, Professor David Rand, professor of information science in the Cornell Ann S. Bowers College of Computing and Information Science. \u00a0https:\/\/business.cornell.edu\/faculty-research\/faculty\/dgr7\/<\/a> \u00a0\u00a0Here is my inquiry:<\/p>\n

To me, what was important was the lesson that fact-heavy conversations seemed to have the most impact on changing a person’s perspective and doing so in a way that endured over time.\u00a0 What I’d like to add [to an essay about \u201cLevers\u201d] is whether this finding can apply to courtroom advocates.\u00a0 In court we don’t have conversations, or said differently we try and be conversational but the jurors can’t speak back.\u00a0 But we still try and win hearts and minds and sometimes have to change attitudes.<\/p>\n

Here is what Professor Rand replied:<\/p>\n

In terms of your question, we haven’t directly tested it, but I see every reason to expect that the same would be true in the context of the courtroom. The key thing is, I believe, that people are paying attention\/engaging with the information being provided (i.e. being in a “high elaboration” state as per Petty & Cacioppo’s Elaboration Likelihood Model). When people are paying attention, facts and evidence are the best way to durably change their attitudes. (When they aren’t really paying attention, you need to resort to other methods).<\/p>\n

What do I take this to mean?\u00a0 Keep on telling stories to shape what jurors remember and how they will organize the facts that emerge across a trial.\u00a0 But use facts – yes, the Aristotelian \u201clogos\u201d approach \u2013 to convince and move those who don\u2019t favor your position to a new belief system.\u00a0 [And, of course, take preventive measures against false facts.]<\/p>\n

And the \u201celaboration likelihood model\u201d that Professor Rand references?\u00a0 You can read about it here – https:\/\/www.simplypsychology.org\/elaboration-likelihood-model.html<\/a> – and there will be follow-up in another Brain Lessons piece.<\/p>\n

………………….<\/p>\n

Professor Rand offered suggested additional readings.<\/p>\n

You might also like these related papers which find the same fact\/evidence-based persuasion in the context of debunking conspiracy theories:<\/p>\n