About the paper
The report is an original research survey about how PR professionals are using, governing and judging generative AI and emerging AI agents in their work.
It is based on a survey of 564 PR professionals fielded from 5 December to 24 December 2025 and distributed primarily via email; responses were reviewed for low-effort patterns and outliers.
The geographic scope is not clearly specified in the report, and the methodology is useful but relatively thin on sampling detail beyond distribution method and data cleaning.
Length: 28 pages
More information / download:
https://muckrack.com/resources/research/state-of-ai-in-pr
Core Insights
1. What is the report’s central argument about the current state of AI in PR?
The core argument is that generative AI has moved from novelty to normal practice in PR, but that adoption has now largely plateaued. The report says AI use has “peaked” at around three-quarters of PR professionals, with 76% already using generative AI in their workflow and only modest shares still undecided or resistant. In other words, the market now appears split between a large majority who have already incorporated AI and a smaller minority who are unlikely to change their minds soon.
The report also argues that this is no longer just an individual experimentation story. Organisations are adapting to AI institutionally: 51% of respondents say their company has an AI use-case policy, and 43% say their workplace offers AI training. That suggests AI in PR is becoming formalised, governed and embedded in workplace practice rather than remaining a purely ad hoc tool used by curious individuals.
At the same time, the report draws a clear line between generative AI and agentic AI. While text-generation and workflow support are mainstream, AI agents have not yet crossed into broad professional adoption. Only 12% say they use AI agents in their work, and most respondents remain uncomfortable with autonomous action without human review. So the report’s bigger message is not “AI is coming”; it is “AI is here, but autonomy is not yet trusted.”
2. How are PR professionals actually using AI, and where do they see the most value?
The report shows that PR professionals are using AI mainly in mainstream knowledge and writing tasks, not in highly specialised or fully automated ways. The most common use case is editing and refinement, cited by 86%, followed by research and insights at 76%, writing and content creation at 74%, and strategy and planning at 68%. Administrative tasks are also significant at 51%, whereas media outreach, measurement and creative asset production remain much less central.
That distribution matters because it shows where AI currently fits the profession best: it acts primarily as a cognitive and editorial assistant. It helps polish drafts, speed up background work, support ideation and assist with planning. The report says PR pros use AI in an average of four distinct work areas, which reinforces the idea that adoption is broad across tasks even if it is not yet deep in every part of the workflow.
Perceived value is also very strong. Eighty-two per cent say AI has improved the quality of their work, while 93% say it helps them complete projects more quickly. When asked where the greatest time savings occur, respondents again point to core communication work: editing and refinement, research and insights, and writing and content creation. This suggests AI’s practical value in PR is currently less about replacing judgement and more about compressing routine labour around producing and shaping communication.
3. What does the report suggest about trust, oversight and the limits of AI in PR practice?
A key theme running through the report is that PR professionals are using AI extensively, but they do not trust it enough to leave it unsupervised. The clearest sign of this is editing behaviour: 98% say they always or often edit AI-generated text before using it. Although the extent of editing has decreased over time, the human review step remains almost universal. That implies AI is accepted as a draft partner, not as a final author.
The report also reveals a selective approach to data input. Large majorities say they avoid entering financial data, personally identifiable information and proprietary or strategic material into AI systems. Yet far fewer avoid entering client or brand names and internal communications. This points to a practical but uneven data-risk mindset: many PR professionals recognise serious information-governance risks, but boundaries around what is safe to share with AI are still inconsistent.
Trust drops even further when the report turns to AI agents. Only 7% say they would be comfortable allowing an AI agent to send messages or publish updates without human review, while about nine in ten are uncomfortable. The main conditions for greater trust are strong evidence of reliability, explicit human approval before publishing or sending, strong privacy and security safeguards, and transparent records of what the agent did. The report therefore implies that the profession is not rejecting automation outright, but it wants clear guardrails, accountability and a human in the loop.
4. What risks, concerns and professional tensions does the report identify?
The biggest concern in the report is not that AI will immediately destroy PR jobs, but that it could erode professional formation. Seventy-seven per cent say the main risk is that younger or newer PR professionals will fail to learn the principles of the profession and rely too heavily on tools. That is a profound concern because it goes to capability-building, judgement and the long-term health of the field.
Other major concerns are also closely tied to quality and professional standards. Sixty-three per cent worry about unscrutinised AI output lowering the quality of communication, while 61% fear content becoming less original or creative and 61% think audiences may be overwhelmed by rising volumes of content. There is also concern that clients or firms may conclude they no longer need content creators. Together, these findings show that respondents fear AI may devalue craft, raise noise levels and weaken the human distinctiveness of communication work.
Among non-users, the resistance is often principled rather than merely practical. The report notes that the 7% who do not plan to explore generative AI often cite environmental impact, plagiarism and broader ethical concerns. Many of them do not appear persuadable: 80% say they do not see themselves using AI in the future no matter what. That is important because it means non-adoption is not simply a training gap for everyone; for some, it reflects a deeper values-based objection.
5. What are the report’s main implications for the future of PR work, capability-building and organisational practice?
The report points towards a future in which the competitive edge in PR will come not from whether someone uses AI, but from how well they use it under proper governance. Since adoption has already stabilised at a high level, the next differentiator is likely to be capability. Respondents say the most important new skills are prompt writing or engineering, ethical decision-making, AI tool evaluation and selection, and data literacy. That suggests the profession is shifting from simple tool familiarity to a broader blend of technical fluency, judgement and governance awareness.
There is also a strong implication for employers. Training and policy appear to matter. Among those open to AI but not yet using it, 60% say training or proven examples would help them start. At the same time, the growth in workplace policies and training since 2024 suggests companies are beginning to respond to that need. So the report implies that organisational maturity around AI is no longer optional; it is becoming part of professional infrastructure.
Finally, the report suggests that PR is heading towards a hybrid future rather than a fully automated one. Paid AI use is rising rapidly, and everyday AI-supported work is now common, but agentic AI remains marginal because trust, accountability and reputational risk still demand human control. The underlying conclusion is that PR professionals are prepared to augment their work with AI, but not to surrender authorship, responsibility or final judgement to it. That is probably the report’s most important strategic implication.