Research onsynthetic authenticity.
Current studies examine how AI earns permission to occupy roles that matter, how people scrutinise it and how institutions govern its authority.
7.7k+
academic records mapped or screened across authenticity and AI-persona research.
13.6m+
public posts, comments, reviews and discussions examined across AI influence, advice and care.
19.1m+
public records and discourse items examined across emerging agentic markets.
A wider programme for AI legitimacy and institutional trust.
The research now follows four questions that institutions need to answer when AI holds a role that matters.
Legitimacy gap opens when an AI can do more than it has been given permission to do. Explore the framework →
Published
International Journal of Human-Computer Interaction
The authenticity paradox.
Saxena, P., & Prahl, A. (2026). The authenticity paradox: The threshold model of synthetic authenticity. International Journal of Human-Computer Interaction. Advance online publication. https://doi.org/10.1080/10447318.2026.2680242
The foundational study. A map of 2,685 articles and 111 AI-persona studies shows where human cues help and where they start breaking trust.
Under review
Human-Machine Communication
When the communicator is code.
Saxena, P., & Prahl, A. (manuscript under review). When the communicator is code. Human-Machine Communication.
People keep re-checking what kind of thing is speaking, and that quiet check shapes what they are prepared to believe or permit.
Under review
International Journal of Advertising
The transparency tax.
Saxena, P., & Prahl, A. (manuscript under review). The transparency tax on virtual influencer engagement. International Journal of Advertising.
In an observational analysis of 1,531 Instagram posts, AI disclosure was associated with 19.2 per cent higher engagement. Lower felt authenticity appeared where disclosure became unusually prominent.
Accepted, in production
Journal of Advertising
Crowd forensics and warrant braiding.
Saxena, P., & Prahl, A. (in press). Crowd forensics and warrant braiding: Collaborative persuasion knowledge in AI influencer threads. Journal of Advertising.
Public comment threads become forensic investigations. One viewer's private suspicion turns into the next viewer's public cue.
Under review
Internet Research
Advice goes public.
Saxena, P., & Prahl, A. (manuscript under review). When advice goes public: Carrier framing and community adjudication of AI and human recommendations. Internet Research.
Across 174 carrier posts and 87 full threads, communities audit AI and human advice for different faults. AI advice is called generic; human advice is questioned on process. The split runs 55.8 standardised percentage points.
Revise & resubmit
Journal of Macromarketing
Burdenless listening.
Saxena, P., & Prahl, A. (manuscript under revise and resubmit). Burdenless listening in AI care markets. Journal of Macromarketing.
Across 3,670 posts and reviews, people described companion AI as easier to confide in because it asked nothing back. Separation could feel like loss.
