Those working in PR and comms may already be at the forefront of using AI, but what do they need to know about how market research is changing too? Over the next few years, new tools and techniques will revolutionise thought leadership research, and PRs need to keep pace to avoid being left behind. AI is reshaping how we collect, analyse, and present insights, creating opportunity and risk in equal measure. From debates around synthetic data to AI-written press releases, the question isn’t whether to use AI in research, but how to use it responsibly without compromising credibility or client reputation.

The role of synthetic data, artificial data designed to mimic real-world data, in research-driven PR campaigns is likely to be limited. While it can help model missing respondents or top up hard-to-reach audiences in survey, it’s unlikely that journalists will publish stories based solely on synthetic datasets, except in clearly experimental contexts.

Where researchers are using AI to add genuine value to PRs is in deeper analysis, stronger story development, and improved data quality.

Some key trends of AI and research together driving forward PR, that I believe will only grow in 2026 include:

AI-driven open-text analysis for newsworthy headlines

AI bringing the benefits of rich qualitative insights into quantitative survey scripts—achieving qual at scale. AI tools can analyse and quantify open-text responses, reducing reliance on rigid questions and allowing respondents to answer in their own words. This makes surveys more flexible, the responses more human, and often more revealing, making for more newsworthy headlines and deeper insights.

Conversational surveys for stronger storytelling

AI is also making surveys more conversational, with capabilities to probe respondents while the survey is live. This reduces missed story angles and captures context that would traditionally require follow-up qualitative work, resulting in data that supports stronger, more compelling storytelling.

Detecting bad respondents and protecting against reputational risk

Another area where AI is already proving invaluable is data quality. Identifying poor-quality respondents has long been a challenge, particularly in large online surveys. AI-driven tools that operate while surveys are live can detect poor-quality respondents through keystroke analysis, response pattern monitoring, and comprehension checks, transforming this process. They help ensure cleaner datasets, reduce fieldwork time, and remove the risk of publishing flawed findings. For PRs, this matters because in an LLM-first world, gaining coverage for research stories in top tier media only becomes more important for establishing trust and credibility.

AI-powered data analysis for catching hidden insights

When it comes to analysis, AI can identify patterns that might otherwise be overlooked. For PR teams, this could mean spotting unexpected trends, a sharp contrast between stated attitudes and actual behaviour, or a subgroup whose views challenge the dominant narrative. These insights can also form the backbone of distinctive media stories.

However, usage of AI tools for data analysis should come with a few caveats. We have already caught some potentially embarrassing examples where analysis has been outsourced to an AI tool, resulting in wildly incorrect statistics and even confident answers to questions that were never asked in the survey. It is a useful reminder that AI can sound convincing while being completely wrong. So, when using AI to analyse data or draft press materials, 

PRs should follow a few core principles:

1. Expertise matters. People with professional knowledge of research are far better equipped to ask the right questions and sense-check the answers an AI tool produces. If you would struggle to perform a task yourself, asking an AI to do it unsupervised is a big risk.
2. Prompt writing matters. If it’s safe to use an AI tool, it’s worth noting that the way a prompt is framed can heavily influence AI output. Just like we do in a market research survey, avoid language that assumes a conclusion and aim for neutral wording that allows the data to speak.
3. Verification is essential. Always test outputs by running tasks across different AI tools where possible to check the reliability of results and never rely on analysis you cannot confidently validate. This is where working closely with your research partner really matters. Every interpretation should be checked before publication.

At Opinium, we’re using AI to transform every stage of the research process and we’re building an AI-enabled agency with people at its core. We’re committed to using AI to enhance human skills and drive innovation for our clients — always guided by responsibility and rigour. We know many of our clients are transforming their PR workflows using AI and we’re doing the same with our researchers. Supplementing our respective skillsets and expertise with AI, we can work together to produce even more impactful thought leadership reports.