About the paper
This sector-specific industry analysis draws on the LexisNexis Future of Work 2026 global survey of 1,400 professionals across more than 20 industries, conducted online through Sago using random sampling from 8–17 September 2025.
The report focuses on PR, media and communications professionals worldwide, but the size and geographic composition of this sector-specific subsample are not clearly specified in the report.
It is best characterised as a secondary sector analysis of original survey research, supplemented by LexisNexis interpretation and recommendations.
Length: 7 pages
More information/download: https://www.lexisnexis.com/en-us/industries/public-relations-communication/ai-pr-comms-report.page
Core Insights
1. What is the report’s central finding about generative AI adoption in PR, media and communications?
The central argument is that PR, media and communications professionals face an unusually difficult combination of high potential value from generative AI and high reputational exposure when it goes wrong.
Content creation is the sector’s leading use case for genAI, reported by 55% of respondents. This makes the technology highly relevant to everyday professional work because producing written and other communicative content is central to the sector. At the same time, adoption is relatively cautious: only 32% say they use genAI frequently or always, compared with 52% across all industries, while 45% remain at an early experimental stage.
LexisNexis interprets this not simply as slow adoption, but as a consequence of the particular risks facing communicators. Errors in many other professions may remain internal, whereas inaccurate, fabricated or misattributed information in PR or media can immediately become public. The report therefore presents the sector as caught between the efficiency benefits of AI and the requirement to protect organisational and professional credibility.
Its broader thesis is that the sector now needs to move from asking what AI can do towards establishing whether AI-generated information can be trusted. The transition is characterised as a shift from AI “capability” towards AI “credibility”.
2. Why does the report consider misinformation and reputation particularly serious risks for this sector?
Misinformation is the standout concern among PR, media and communications professionals. Some 59% identify misinformation as a leading concern, compared with 40% across the full cross-industry sample.
The report connects this anxiety directly with the nature of communications work. Published material can be scrutinised by journalists, customers, clients and wider public audiences, and errors can become permanent parts of the public record. In media particularly, credibility is based heavily on accuracy, so hallucinations, incorrect sourcing or misattribution can undermine trust in both the individual communicator and the organisation.
There is also a practical verification problem. The report says that 36% struggle to determine whether AI outputs are based on high-quality sources, while 32% cannot identify hallucinations or incorrect outputs. These weaknesses are especially consequential when AI is used to produce press releases, articles, features and other externally published materials.
Consequently, PR and communications professionals appear to place greater value than many other sectors on mechanisms that make AI output auditable. Transparent data sources are identified by 45% as the leading way to increase trust, while 41% say human oversight increases trust.
The report therefore treats misinformation not merely as an AI accuracy problem but as a reputation-management problem. For communications functions, uncertainty about where information comes from and whether it can be verified becomes an organisational risk.
3. How serious is the gap between AI use and AI competence in PR, media and communications?
The report portrays the competence gap as one of the sector’s most important weaknesses.
Only 50% of PR, media and communications professionals describe themselves as very or extremely confident in their ability to use genAI, compared with 64% across industries generally. More strikingly, only 5% say they can confidently explain large language models, compared with 29% across the overall sample.
The report therefore distinguishes between being able to operate an AI tool and understanding enough about the technology to assess its limitations. This distinction becomes critical when professionals need to judge source quality, identify hallucinations or decide whether an AI-generated claim is reliable enough to publish.
Training alone does not appear to have solved the problem. Across the entire survey, 82% of professionals receive some form of AI training, yet 41% of PR, media and communications respondents still cite lack of training as a leading barrier. LexisNexis consequently argues that organisations need to move beyond basic AI awareness towards specialised training in verification, source assessment and recognition of situations requiring human checking.
The report’s underlying concern is therefore not simply that communications professionals need more experience with AI. They need a different kind of competence: the ability to evaluate AI-generated information safely in contexts where mistakes can create public consequences.
4. What contradiction does the report identify between cautious AI adoption and unauthorised AI use?
One of the most interesting patterns is that relatively cautious overall adoption coexists with substantial unauthorised use.
Only 32% of PR, media and communications professionals report using genAI frequently or always, suggesting comparatively restrained adoption. Yet the report also says that large numbers use AI tools without organisational approval.
There is an internal numerical inconsistency in the report on this point. The narrative states that 63% engage in unauthorised usage compared with 53% overall, while the accompanying chart gives the PR, media and communications figure as 59%. The report does not explain the discrepancy, so the exact sector figure cannot be established confidently from the document.
Regardless of whether the figure is 59% or 63%, the underlying pattern is clear: formal adoption may be cautious while informal or unsanctioned use is widespread. Among unauthorised users, 62% reportedly say deadlines will not wait, while 61% point to the need for efficiency.
This creates what the report effectively presents as a governance paradox. Reputational anxiety makes professionals wary of AI, but everyday workload and deadline pressures simultaneously encourage them to use it outside approved systems.
The implication is that restrictive policies alone may not control AI use. If employees perceive AI as necessary to complete their work efficiently, banning or delaying access can push usage outside organisational governance rather than eliminating it.
5. What does the report recommend leaders should do, and what perspective underlies those recommendations?
LexisNexis proposes five main actions. Organisations should provide purpose-built AI tools for high-stakes content work; invest in specialised training focused on quality verification; embed human validation checkpoints into AI-supported workflows; create reputation-protection frameworks defining permitted and prohibited use cases; and accelerate the transition from experimentation towards properly governed integration.
These recommendations follow logically from the report’s diagnosis. If source verification is difficult, tools should make sourcing more transparent. If users struggle to detect hallucinations, training should concentrate on validation rather than simply prompting or basic AI literacy. If deadline pressures encourage unauthorised adoption, organisations need sanctioned tools that employees can realistically use in everyday workflows.
The report also argues that human oversight should be designed into processes rather than treated as an abstract principle. Suggested measures include spot-checking AI-generated material and mandatory human review for particularly consequential outputs.
At the same time, the recommendations should be understood within the report’s commercial perspective. LexisNexis is not an independent academic institution but a provider of information and AI-enabled services. The report explicitly recommends purpose-built tools, reputable technology vendors and transparent source environments, and concludes by inviting readers to explore LexisNexis AI solutions and its content resources.
The report therefore serves two purposes simultaneously: it presents survey-based evidence about genuine challenges surrounding AI adoption in communications, while also framing those challenges in ways that support LexisNexis’s proposition around trusted data, verifiable sources and enterprise AI tools.