Blog

Trust, tools and the human touch: Our view on AI in PR

How we’re making AI work for better communications

August 2025

A new chapter for content and communications

Generative AI is making research and writing faster but it’s also making it harder to know what is original, accurate and created by people - not machines. As tools like ChatGPT become part of everyday workflows, marketers and PR professionals are working out how to use them responsibly. That means thinking not just about speed or scale but about clarity, trust and how we show real expertise.

New concepts like Generative Engine Optimisation (GEO) are shaping how we think about content that performs well in search and holds up to scrutiny. At its best, GEO blends the technical with the credible. That is where thought leadership plays a key role. Original opinions, trusted voices and experienced perspectives are exactly what search engines and readers are looking for.

This is also where content like blogs, FAQs and even press releases now hold new value. While press releases are becoming less impactful as direct media tools, they still contribute to visibility and help organisations show consistency across platforms. It is another reason why messaging needs to be both structured and grounded in expertise. But there’s still a long way to go in understanding what good looks like.

We don’t ask AI to think for us. We only use it to help shape what we’re already thinking.

Balancing speed and substance

We’ve been exploring the role of AI in our own work and in conversations with peers and partners. As a team, we’re using tools to support research, improve workflows and generate ideas but we’re clear about where we draw the line. Final messaging, tone of voice and client-facing outputs like insight pieces, thought leadership, media commentary and strategic positioning are never handed over to AI. That still requires human clarity, editorial judgment and context.

This came through clearly in PR Moment’s recent AI in PR masterclass. Journalists are using tools like Perplexity and Notebook LLM to surface new voices, test angles and check tone. But they’re becoming more selective about who they trust. Concerns around misinformation and the rise of AI-generated spam are making it harder to earn attention especially in media relations.

The same session showed that while AI can support drafting structure and summaries, it still can’t replace the insight or experience that makes content useful. A story built entirely by a tool might look polished, but it often lacks what makes a message credible or meaningful.

Getting better results with better prompts

Crucially, the way you frame a question or brief can completely shape the quality of the output and perfecting this part of the process takes time and experimentation. A well-structured and clearly defined prompt makes a big difference in the quality of AI outputs.

We’re continuously refining how we write prompts and how we train GPT models to align with Hattrick’s tone, clarity and voice. That means using examples, embedding brand context and making sure the tools reflect how we actually speak and write. It’s not about achieving polished results but about building helpful and repeatable processes that save time without compromising quality.

Keeping sustainability in the frame

As we test new tools and approaches, we’ve also been thinking about the bigger picture. One conversation that’s helped shape our thinking is The AI Paradox: Powering Progress or Draining the Planet, part of the Curiosity, Applied podcast by Verdantix. The episode explores AI’s potential to support sustainability from optimising energy use to improving climate modelling. But it also highlights the growing environmental cost of AI’s rapid adoption and the pressure this puts on infrastructure and energy supply.

For us, it’s a timely reminder that responsible AI use isn’t just about accuracy and tone. It’s also about understanding the systems we’re working in and being thoughtful about the impact our tools can have - not just on output but on the planet. In sustainability communications, this kind of thinking matters and it is not just what we say but how we work behind the scenes that builds credibility.

How we’re using AI at Hattrick

At Hattrick, we’ve chosen to use AI in a way that supports our processes, which means:

Using it for early-stage research or idea development
Using it to help us stay aligned with our Green Claims Policy by tracking evolving sustainability terms and definitions
Training models to better reflect our tone, structure and writing style
Exploring structure or tone options, especially in content planning
Never publishing content without full human review and editing
Proof reading any AI-sourced data and links

We’ve also seen first-hand where it goes wrong. Placeholder results, made-up stats or non-existent links are still common. So while it’s a helpful tool, it’s only one part of how we work. We don’t ask AI to think for us, we only use it to help shape what we’re already thinking.

AI as a tool in PR and marketing

AI use is growing across the industry but confidence still varies. According to a 2024 report from CoSchedule, 85% of marketers now use AI tools to support content creation, with 83% saying it’s helped them work more efficiently.

However, trust is still a concern. As per Marketers Media, 72% of journalists say they worry about factual errors in AI-generated content. And with more teams relying on tools to fill content gaps, that scrutiny is only likely to grow.

This is a reminder that while these tools are helpful, they’re not a replacement for experience or editorial quality. As PR professionals, we still need to ask: does this reflect the brand, is this genuinely newsworthy and does it help build authority and trust?

The challenge ahead

AI will keep evolving and it will likely become a standard part of how agencies and in-house teams work. But for purpose-led brands, especially in sectors where trust matters, the challenge isn’t just keeping up - it is staying thoughtful.

We’re choosing to focus on content that’s clear, credible and grounded in real insight. AI can help shape that but it can’t replace the value of good judgment, shared context or a strong point of view.

Moving forward with clarity

If you're looking to bring more efficiency into your content strategy, AI can help. But it works best when paired with original thinking, brand experience and a strong editorial lens. A good blend of content skills and AI tools is the need of the hour. We believe the future of PR and content will be shaped by teams who use AI to support their ideas, not define them.

📥 Explore our approach and work with us