 |
| Nathan Richter |
The use of AI-generated “synthetic” research is spreading through the public relations industry. Research firms are promising faster turnarounds and lower costs by replacing real survey respondents with large language models to simulate how audiences might respond.
It makes sense why a fast-paced industry like communications would see the appeal. Budgets for research are shrinking and AI is becoming commonplace throughout workflows. However, there is a key distinction between research built from predicting likely answers based on known data and research that measures real human thoughts and feelings.
This distinction matters for organizations that are looking to frame complex data into compelling stories and drive thought leadership.
Synthetic Audiences Are Not Real Audiences
At its core, these AI-generated models are created using machine learning to mimic human responses based on datasets of consumer behavior.
For general population consumer studies focused on established, well-known categories, this can be a tempting step for basic language learnings early in the creative process. But it is crucial to understand the limitations of this type of modeling.
You’re not getting real people from your target audience giving you their authentic responses. You’re getting AI-generated synthetic personas created based on patterned behavior.
At the end of the day, human decision-making is complex and includes both emotional and irrational factors, which synthetic systems can’t account for.
The Risk of Repackaging Old Insights
Another issue that should concern communications professionals is that synthetic data relies heavily on historical data. In this situation, outdated insights are repackaged instead of working strategically to uncover new findings or explore how audiences’ opinions have changed. In other words, it essentially eliminates the newsworthiness of the research.
Based on our experience partnering with some of the world’s leading brands and agencies to achieve top-tier earned media via research, we understand that the most valuable data comes from reactions that challenge long-standing assumptions, illustrate shifting attitudes, or uncover an unexpected conclusion. These findings give PR professionals a unique news hook and the type of data points that not only capture journalists’ attention but encourage top-tier media to cite research.
That becomes harder, if not impossible, to do when the output is pulling from the same patterns and assumptions that already exist.
Transparency Remains Critical
An underrated yet critical consequence is the lack of transparency. Most synthetic research platforms provide little visibility into the data being pulled, assumptions made, or the modeling frameworks shaping the research outputs. This often leaves communication teams in the dark on what information was used to train the system, how audiences are being represented, and what biases may exist within the model itself.
This can become a bigger problem when research is used to frame stories that support thought leadership efforts, earned media, and marketing campaigns. Journalists rightfully expect full transparency on where the findings came from, how the research was conducted, and whether the data holds up under scrutiny.
Without confirmation of a sound methodology, brands risk losing credibility amongst consumers and peers, in addition to jeopardizing important relationships with media. At a time when journalists are battling misinformation, it remains more imperative than ever for PR professionals to provide verified research and transparency.
Where AI Fits in Modern Research
While these issues should grab your attention, it doesn’t mean AI doesn’t have a place in modern research. Methods like synthetic research can still be useful as a starting point for testing language in a preliminary draft. But it should only play a supporting role within the research process.
The most important thing to remember is that research is only valuable if it’s trustworthy.
Today’s journalists want findings they can rely on and data that reflects what real audiences think in an era of misinformation. In fact, Cision’s 2026 State of the Media report revealed that 47% of journalists want PR professionals to send more data or research, which is the top requested resource, and 33% said credible data or research makes them more likely to engage with a pitch.
AI will continue to be part of research workflows, but accountability still belongs to the PR professionals collaborating with research partners to generate unique insights on behalf of brands. Communications teams should lead with data and confidence, validate key findings, understand the methodology being used, and stand behind the conclusions shared publicly.
That is what ultimately gives research value in the first place.
***
Nathan Richter is a senior partner at Wakefield Research.
AI-driven Metrics, algorithms, and data optimisation have become the latest obsession for many, and the true fundamentals of B2B marketing have been forgotten.
In a recent “PR’s Top Pros Talk…” podcast episode, Will Reese, Chief Innovation Officer at Inizio Evoke, and Theresa Dolge, Chief Media Relations Officer at Inizio Evoke, joined Doug Simon, CEO of D S Simon Media, to discuss the growing influence of AI on earned media, local journalism, and public relations.
Research by DBC and the 4As shows that by replacing entry-level positions with AI, ad agencies are inadvertently eroding the apprenticeship model that historically has served them well and increasing the probability of a "hollowed-out" middle management layer.
Fresh off receiving her PhD at the University of Missouri School of Journalism, Lea-Ann Germinder has introduced AELHA™, an original model for keeping humans at the center of AI adoption.
The inaugural list maps the figures — most of whom are not in the PR profession — whose decisions now influence whether the work the profession produces gets surfaced when AI engines answer questions about a category.
