The AI reality gap in corporate communications and public relations
There are several reasons why the individual effectiveness and efficiency benefits of AI in corporate communications and public relations haven’t yet scaled to teams.
Industry surveys tell the story of the wholesale adoption of AI. In February, Cision and PRWeek reported that 92% of practitioners use AI, and 32% use it frequently (Global Comms Report 2024). USC Annenberg reported early this month that the number of frequent users had increased to 66% (AI Activated).
It simply isn’t the case. The reality is that AI hype is giving way to disillusionment as organisations figure out the best way to use AI to benefit teams.
AI adoption reality versus hype
It’s two years since the launch of the ChatGPT large language model by OpenAI. Claude from Anthropic and Gemini from Google quickly followed. Professional versions of any of these applications are available to anyone with access to a browser for less than £20 per month.
This accessibility has enabled individual practitioners to understand the potential of AI in management and functional communication roles. AI promises increased cognitive function, efficiency, and effectiveness.
There’s a huge third-party market of shiny tools that will help practitioners with every aspect of practice. A 2023 study by Andrew Bruce Smith and me for the CIPR discovered more than 10,000 (AI tools and their impact on public relations practice).
AI is equally capable of supporting practice with the management tasks of building stakeholder plans, horizon scanning, and red teaming, as well as the tactical tasks of administration, editing or summarisation, and content creation.
It will work well to support a practitioner one-on-one as a professional assistant or mentor with all these tasks and more, representing incredible value.
The promise of democratising knowledge and skills can only be good for practice. Although it will inevitably impact jobs, no one can say how soon.
Individual benefits versus team implementation
The challenge is that AI technology has yet to be applied equitably in practice. It works well as an expert administrator to support individual practitioners, but workflow doesn’t exist to support teams.
Bias is also an issue. AI systems are coded to be optimistic and progressive. You need to work hard to force them to work critically.
Initial studies, such as the landmark Boston Consulting Group study called Navigating the Jagged Frontier published by Harvard Business Review in 2023, suggested AI would deliver 40% effectiveness and 25% efficiency benefit (Navigating the jagged technological frontier: Field experimental evidence of the effects of ai on knowledge worker productivity and quality).
The recent USC Annenberg study reported that 70% of practitioners believed AI helped them produce better work, while 73% say it allows them to work more quickly.
Herein lies the issue. AI in public relations practices can deliver significant benefits for individuals, but that benefit has not yet been demonstrated to scale well for teams.
The initial energy and heat in the market created by large language model vendors driven by venture capital investment is cooling as practitioners face the cultural issues of embedding AI into management and communication workflow.
The workflow challenge
USC Annenberg says it’s a training issue. That’s almost certainly true, but it’s overly simplistic. There’s a bigger issue.
There isn’t a standard operating model for public relations practice. It is contextual to an organisation, its audiences or public, its market and its structure. Every agency or communication team must break down its processes and apply an innovation model.
It’s challenging and hard work.
Emergent workflow solutions
Vendors have recognised the issue. Microsoft quickly embedded a version of ChatGPT in its office application called Copilot. Google followed and has done the same with Gemini. However, the integrations have yet to bring the full benefit of either application to the enterprise.
Customised large language models are a potential route. Here, ChatGPT is trained on an organisation’s body of knowledge, such as corporate documents and web content. It enables practitioners to query historical content and ensure they follow corporate style.
Third-party tools have also recognised the need to embed AI within existing workflows. Vendors such as Otter and Jasper are overlaying knowledge bases, project management, sharing, and version control.
The path forward
The promise and potential of AI in corporate communication and public relations and communication is clear.
Individual practitioners are already realising significant benefits from large language models in both management and functional roles. Early studies and practitioner feedback demonstrate improvements in both efficiency and effectiveness.
However, the wholesale adoption claimed by industry surveys masks a more complex reality.
The challenge lies not in the technology itself, but in scaling these individual benefits across teams and organisations. Without standard operating models and proper workflow integration, corporate communications and public relations teams are struggling to embed AI systematically into their practice.
The route forward requires organisations to tackle this challenge head-on: breaking down their processes, developing appropriate workflows, and implementing careful change management.
Vendors are beginning to address these needs through enterprise solutions and workflow tools, success will ultimately depend on each organisation's ability to adapt and integrate AI in a way that fits their specific context and needs.
As the initial hype around AI gives way to practical implementation, the focus must shift from individual tool adoption to organisational transformation. Only then can the public relations industry fully realise AI's potential to enhance individual and team performance.
Further information
Wadds Inc. published a management briefing paper called The use of AI in corporate communications and public relations: the story so far in June 2024 highlighting the challenge of integrating AI into workflows while addressing governance issues and striving for increased efficiency and effectiveness.
If you need help with any of the issues raised in this blog post please contact Stephen Waddington.
