AI in PR 2025: Perception Outpaces Practice as Agencies Pull Ahead

2025 AI in PR Survey: Perception Outruns Practice

AI has moved from trials to everyday work. The first PRWeek and Boston University AI in PR survey captures where teams stand and what's holding them back.

Fielded February-May 2025 with 719 in-house and agency respondents, the survey measured agreement on a five-point scale. The headline: organizations say they are pro-innovation, yet actual AI use is still light.

The perception gap

Pros rate their employers highly for innovation (4.33) and collaboration (4.22). But personal AI use at work sits at 2.80.

Many also believe peers use AI more than they do (3.46). That gap breeds quiet anxiety - and a clear call to turn interest into execution.

Why adoption lags: readiness and leadership

Infrastructure is thin (2.64), especially in-house. Pros want to move beyond experiments, but they lack the systems, guidance and resources to do it well.

Leaders confirm the pattern seen with past tech shifts: readiness trails desire. The difference now is speed - institutions are catching up faster, but the gap still exists.

In-house vs. agency: two different realities

Limited access to AI experts (2.81) fuels a leadership void, felt most by in-house teams. Compared with agencies, in-house respondents feel less heard on AI (2.88 vs. 3.66), less understood on adoption challenges (2.86 vs. 3.59), less informed in time (2.79 vs. 3.82) and face tighter resource access (2.90 vs. 3.14).

Leaders need to set the tone: clear guardrails for public tools now, with a path to secure internal platforms later. Perception of PR as a cost center still limits attention, investment and sponsorship from the C-suite.

Agencies pull ahead

Agencies spent early cycles on security, IP and governance. That groundwork now pays off. Teams report strong benefits in automating monitoring (4.00) and analyzing large datasets (4.14), with AI becoming a new axis of differentiation.

Some groups secured enterprise-grade LLMs for confidentiality and content rights and built proprietary insights engines that feed smarter pitching and planning. Others use holding-company platforms to spin up reporter agents, mapping tone, likely questions and blind spots in minutes - and showing up to interviews better prepared.

How leading sectors use AI in comms

Tech, financial services and healthcare are ahead because other functions paved the way. Banks, for example, have used AI to tear through complex documents for years and now extend LLM access to comms teams.

Practical use cases are clear: first-draft messaging, summaries of long materials, content calendar ideation, quick explainers for complex topics and anticipating stakeholder questions. AI accelerates the work; teams still own the judgment.

What AI can't replace

Concern about job loss is moderate (employment 3.27; career advancement 3.15). Entry-level tasks are most exposed as agentic AI takes on repetitive work, and some teams are already adjusting hiring at the base layer.

Still, AI isn't compatible with everything in comms (2.73). Emotional intelligence - reading a room, aligning people, landing messages - remains human territory.

The risk: thinking and creativity decay

Top concern: reduced critical thinking (4.13), with creativity close behind (4.00). Overreliance on AI can dull the skills senior work requires.

This is a solvable problem. The organizations that put policies, workflows and training in place now will keep speed without losing rigor.

AI prep playbook for PR leaders and teams

• Audit readiness: Map your current infrastructure, access, security, data policies and content rights. Close the basics first.
• Pick 3 use cases: Media monitoring, insights from large datasets and first-draft content are reliable starters. Define KPIs for speed, quality and accuracy.
• Set guardrails: No sensitive data in public tools. Require source logging, citation, fact-checking and bias checks. Align with frameworks such as the NIST AI Risk Management Framework.
• Secure your stack: Prefer enterprise LLMs with encryption, admin controls and content rights. Build a private workspace and reusable prompt libraries. Pilot agents for media research and Q&A drafting.
• Upskill the team: Train on prompting, verification and human-in-the-loop review. Appoint AI champions inside each team. For structured learning by role, see Complete AI Training: Courses by Job.
• Redesign workflows: Require human brief → AI assist → human edit → legal/SME review. Protect original ideation time before AI use to preserve thinking and voice.
• Measure and report: Track time saved, quality gains and error rates. Share wins with the C-suite to move PR from cost center to capability.
• Evolve talent strategy: Retrain for analysis, creativity and client counsel. Rethink entry-level roles and create apprenticeships that build judgment, not just task execution.
• Create governance: Form an AI council, set policy, review vendors quarterly and maintain an incident response plan for data or content issues.

Quick benchmarks from the survey

• Employer receptivity to innovation: 4.33
• Cross-team collaboration: 4.22
• Personal AI use at work: 2.80
• Perceived peer AI use: 3.46
• Organizational infrastructure: 2.64
• Access to AI experts: 2.81
• Top benefit - automate monitoring: 4.00
• Top benefit - analyze large datasets: 4.14
• Concern - employment impact: 3.27
• Concern - career advancement impact: 3.15
• AI fit across all comms tasks: 2.73
• Risk - reduced critical thinking: 4.13
• Risk - reduced creativity: 4.00

Bottom line

Perception is ahead of practice. Agencies that invested in foundations are reaping returns; in-house teams need clearer sponsorship and infrastructure to catch up.

Adopt AI with intent: lock down the basics, standardize a few high-value use cases, protect thinking time and measure results. That's how PR teams keep speed and raise their value at the same time.