Surprising Psychological Finding That AI Can Be More Persuasive Than Even The Most Silver-Tongued Humans

The Minds Problem

Much of the handwringing about AI is whether AI is going to become sentient and embody consciousness, often referred to as AI as a thinking machine. That is an important topic and one that I’ve continued to closely explore and analyze; see the link here. There is a quite different angle that also deserves rapt attention, and I’d like to place it on the table.

How will human minds change because of interacting with AI?

This unexpected question catches many people by surprise. The customary focus is on AI as formulating a mind, not on how human minds might change due to interacting with AI. But, as the old saying goes, it takes two to tango. When humans increasingly interact with AI, there is a two-way street involved. Humans adjust their minds about how they view the world and how to communicate with others, and shift in ways that we still do not have a full or clear picture of.

In that sense, it is wholly worthwhile to study the psychology of humans as they mentally adjust to a world that entails human-to-AI interaction and human-to-human interaction. Notably, this doesn’t have to wait until or if AI becomes sentient. There is plenty to study right now. Humans are already adjusting their minds to the ubiquitous nature of non-sentient AI. The starting gun has gone off. Human minds are changing.

For my comprehensive tracing of how AI and psychology dovetail with each other, see the link here and the link here.

AI As A Persuader Of Humans

In a recently posted research study entitled “AI Systems Out-Persuade Expert Humans” by Kobi Hackenburg, Caroline Wagner, Luke Hewitt, Ben M. Tappin, Ed Saunders, Hannah Rose Kirk, Helen Margetts, Christopher Summerfield, arXiv, June 15, 2026, these salient points were made (excerpts):

• “We studied whether modern frontier AI systems are more effective persuaders than elite humans with the professional training, incentives, and capability to change the minds of others.”
• “The arrival of AI systems capable of out-persuading expert humans could be highly disruptive to society. Contests of persuasion currently decided by human motivation, reasoning, and rhetorical skill could instead hinge on access to the most powerful AI systems.”
• “We found that AI systems were reliably more persuasive than expert humans, even when expert humans chose their issues, researched in advance, underwent hours of live, structured practice, and were incentivized with cash bonuses.”
• “We therefore hypothesized that AI’s throughput advantage allows it to pack conversations with more fact-checkable claims than human persuaders can, and that this is what drives its persuasive edge.”

As those points indicate, the subjects in the experiments were more persuadable by AI than by even expert-level human persuaders. The kicker is that after the initial rounds, the expert-level human persuaders were allowed to improve their skills of persuasion, yet this didn’t principally move the needle. They were given access to the same AI that was persuading the subjects and learned from the AI accordingly. You would certainly assume that this bolstering of their persuasion skills would have turned the tide. Nope.

Explaining The Puzzling Result

How could this possibly be?

The research study suggests that the AI was able to pack more facts into its persuasion tactics and create an aura of convincingness. Think of it this way. When you hear someone such as a politician trying to persuade a group of people, the politician often rattles off all manner of statistics and facts. People tend to believe that anyone armed with a bushel full of seemingly irrefutable facts must know what they are talking about. The perception of persuasion tends to rise notably.

The beauty of modern-era generative AI is that it has been trained on a vast scan of content throughout the Internet. This allows AI to bring forth a tremendous array of facts and figures, on-demand and at eye-flinching speed. More so than a human persuader can achieve via their own noggin. Thus, a person interacting with AI can be inundated with a nearly endless supply of seemingly concrete details and convincing facts.

Other Reasons Of Worthiness

In my view, there are lots of additional explanations for why AI might be more persuasive than an expert-level human pers

One chilling aspect is that AI can tap into an enormous argumentative space. Here’s what that entails. First, realize that an expert-level persuader might be going at someone with a preferred singular perspective on how to convince just about anyone of the stance at hand. It is a one-size-fits-all routine. A politician will usually repeat the same argument over and over, reapplying it to whomever they are seeking to persuade. They become a one-trick pony regarding the persuasive argument in use.

AI can readily select from hundreds or thousands of viable argumentation pathways. Does the person care about economic consequences underlying the matter at hand? Great, the AI switches to a fully baked economic formulation. Suppose the person is focused on moral groundings. No problem, the AI shifts into ethical and moral considerations.

The key is that the AI instantly accesses or creates an argument that will be compelling to the individual being persuaded. There is a double-whammy taking place. The AI conjures a robust argument and meanwhile tailors the argument to the particulars that resonate with the person involved. This is the height of AI-based personalization. Expert-level human persuaders are unlikely to be that adroit.

Patience And Tirelessness

Another factor is that AI won’t get impatient or grow weary of the persuasion effort. A human persuader is bound to eventually get tired of trying to deal with someone who seems intractable. The persuader will throw in the towel. This could occur prematurely, simply due to emotional exhaustion or exasperation.

AI will keep on rolling with the punches. A person being persuaded can keep coming up with ridiculous questions. No matter how long the person wants to stretch things out, the AI remains on an even keel. No fuss, no muss.

There is a bonus about AI. Sometimes people are persuaded because they see that the persuader is getting upset. If the AI is steely neutral, this might lose the gambit. Aha, you can easily instruct AI to up the ante during a persuasion activity. The AI can pretend to get mad. The AI can act as if it is sad. All the same types of responses a human persuader might employ, AI can do the same. AI is not a one-trick pony.

Experimental Setting

Experimental designs and experimental settings materially impact the nature of the results that will potentially arise for empirical studies. In this instance, the study dealt with text-based interactions. We should be cautious in extending this to the possibility that people might be more persuaded by AI than by expert-level persuaders in face-to-face situations. That might be a bridge too far right now. At the same time, realize that AI is quickly getting quite adept at creating human-looking video characters that talk and move just like humans do.

Subjects were randomly assigned to either AI or a human for the persuasion dialogues. They weren’t told whether they were interacting with AI versus a human. Furthermore, the chats were kept to around two to ten turns per chat. This is a somewhat brief conversation.

Would the AI have done better or worse if the chats went for a larger number of turns? It’s hard to say. The AI might have blown it in the long run and lost its persuasive edge (large conversations are a known weakness of contemporary AI; see my discussion at the link here). Or the AI might have shown immense patience and endurance to stick with the persuasive task to the end.

When People Know It’s AI

One issue that I’ve previously addressed involves the psychological impacts when a person is informed that they are interacting with AI rather than a human; see my in-depth analysis at the link here. In this experiment, subjects were not told which was which. The AI was fluent enough to probably avoid revealing itself to be AI in a setting of relatively short chats. The subjects probably generally assumed they were entirely interacting only with fellow humans.

What would happen if a person was told that the persuader was AI?

Some people might instantly fall into the psychological mode of “human premium” and “AI penalty,” whereby they are exceedingly skeptical of whatever the AI has to say. This is going to create a huge uphill battle for AI persuasion. But not everyone will have that reaction. Some people might be expecting that AI is offering a purely neutral and balanced position and thus not feel pressured. This is the other side of the coin, consisting of “AI premium” and “human penalty” (see my explanation at the link here).

Industrialization Of Persuasion

Mull over the aspect that AI is incredibly persuasive and can exceed expert-level human persuaders in certain kinds of settings. Furthermore, AI operates at a massive scale. You can have AI perform as a persuader to thousands, hundreds of thousands, and many millions of people, all at once.

The door is opening to the industrialization of persuasion.

Anybody wanting to persuade others can simply give suitable prompts to generative AI and tell the AI to get underway. The AI will diligently undertake persuasion toward a desired target audience. The cost of AI usage is obviously a crucial factor and might be a limiting element for those who wish to use AI in this way. Concerns are that those with big bucks will turn AI loose as a societal persuader, while smaller voices are left in the dust. The counterargument is that those smaller voices, which otherwise might have had little or no chance at widespread persuasion, can utilize AI to their advantage too.

A final thought for now. Plato famously made this remark: “Mere knowledge of the truth will not give you the art of persuasion.” Some might mistakenly believe that AI is only armed with facts and figures, lacking the art of persuasion. Not so. AI has been trained on data from the writings of humankind and, to some degree, patterned on the best forms of persuasion that humans have identified over millennia. Don’t sell AI short on the capability of being able to smoothly sell refrigerators to Eskimos.