Current AIs out-persuade professionals in lab settings but (probably) not in the world
This article was created by Forethought. See all of our research on our website.How persuasive are present-day AI systems?Unfortunately, it is surprisingly difficult to know where AI capabilities in persuasion are today (September 2026), or even three months ago.
This article explores the current capabilities of AI systems in persuasion, comparing their effectiveness to human professionals in controlled and real-world settings.
What happened
The article discusses the effectiveness of current AI systems in persuading humans, citing experiments where AIs outperformed human professionals in controlled settings. However, real-world applications and broader impacts remain less clear.
Why it matters
Understanding AI's persuasive capabilities is crucial for developers and professionals in fields like marketing, diplomacy, and education, as it affects how AI tools are utilized and regulated.
- AI persuasion capabilities are strong in controlled environments but less evident in real-world applications.
- Ethical concerns and the need for regulation in AI persuasion are significant.
- Further research is needed to bridge the gap between lab results and practical use.
The bigger picture
The article suggests that while AI shows promise in persuasion, its real-world impact is not yet fully realized, indicating a need for further research and ethical considerations.
The article concludes that AI systems are currently at a level between median and professional human persuasion, with significant potential but also challenges in real-world implementation.
Technical details
- Attitudinal persuasion experiments with AI systems like GPT-4o and Claude Opus 4.1.
- Use of benchmarks such as the Anthropic persuasion benchmark.
- Experiments involving real-world behavior changes, like donation fundraising.
FAQ
How do AI systems compare to humans in persuasion experiments?
In controlled lab settings, AI systems often outperform human professionals in attitudinal persuasion tasks.
What are the real-world applications of AI persuasion?
Real-world evidence is limited, with some examples like AI-generated content and Spiralism, but broader impacts are not yet clear.
What are the ethical concerns regarding AI persuasion?
The article highlights the potential for unethical use of AI in persuasion, including deception and manipulation, which require careful regulation.
This summary and analysis were generated with AI from the original article and may contain errors. Read the original for the full story.
Read the full story
Originally published by LessWrong on Oct 5, 2026.
Comments
Sign in to join the discussion.