How to appropriately use AI in media relations

AI tools are becoming commonplace in media relations, but concerns about authenticity are raising red flags

Qwoted, a network connecting media with brands, experts and small businesses added a “Flag for AI” option for journalists shortly after ChatGPT launched. In late 2024, the company incorporated a tool to check pitches for AI-generated content.

But AI doesn’t have to be a dirty word, said Thekla Eftychiadou of FINN Partners. Her agency uses AI in earned media depending on the client’s comfort level.

“You can’t remove the ‘relationship’ from ‘media relations,’” said Eftychiadou, a senior partner in FINN’s Global Tech practice and co-lead of its AI Working Group.

“Our policy is: use it as a helper, as a superpower,” she added. “That human element is the most important aspect.”

Making AI a PR pro’s ‘superpower’

At FINN Partners, the media relations team uses AI in four main areas: data analysis, story ideation, conversation transcription, and identifying reporters and outlets to pitch.

For data analysis, AI helps FINN quickly surface patterns from complex information, such as survey results, social media data or corporate earnings reports. Spotting these trends early can inform proactive story development. For example, a spike in customer sentiment around a new product feature might prompt a timely media pitch or prepare the C-suite for questions.

AI also supports story ideation by generating early ideas and angles. Typing in a broad prompt like “trends in electric vehicles” might uncover lesser-known storylines, such as regulatory changes or evolving consumer habits, that inspire new pitches.

“The way we used to browse news sites or use Google, we’re now also using AI tools to get ideas, story angles and spark initial thoughts,” Eftychiadou said.

Though FINN has experimented with AI to write draft media pitches, they tend to keep it to outlines or streamlining thoughts about approaches.

“It’s not going to do as good a job as a communicator… if they know their story,” Eftychiadou said. “The success of media relations has to do with knowing your audience, knowing what the reporter really cares about.”

One of the most effective uses of AI, Eftychiadou finds, is transcribing subject matter interviews – whether for media training or a formal interview. Tools like Otter record and summarize a conversation, allowing the media team to stay engaged during conversations rather than having to take notes. These recordings and transcribed interviews, all verified after the fact by humans, are useful for things like crafting FAQs or talking points.

FINN is also exploring how AI can help select reporters and outlets for particular stories using what Eftychiadou refers to as content resonance or resonance prediction. Relationship-building is still at the heart of media relations; however, some tools can now suggest journalists likely to be interested in specific topics, Eftychiadou said. She stressed a belief that “this tech will improve over time, even if it’s not quite there yet,” noting limitations in terms of its scope and accuracy.

Brands or firms could also consider using AI to build digital personas of specific reporters or types of publications, whether it’s a trade publication or a talk-show-style podcast.

Keeping things human

While Eftychiadou knows AI will be central to PR’s future, she cautions against overreliance. When tech runs too much of the process, it becomes noticeable.

“Are they more suspicious? Absolutely,” Eftychiadou said of reporters. “They’re getting thousands of media requests. With the efficiency AI brings, they’re getting more.”

Tino Bovenzi, a TV reporter for WJW/Fox 8 in Cleveland, said he receives more pitches than ever. He skims for potential stories but is “bombarded with pitches that are just so misplaced, relating nowhere close to our coverage area.”

Mistakes like addressing emails to the wrong person or using phrasing that “doesn’t feel human” make him suspect AI involvement. That isn’t necessarily a dealbreaker, he said, as “a good pitch is a good pitch,” but “quality over quantity” is important.

One way to address reporter suspicion, according to Qwoted co-founder Matthew Kneller, is to add a human element to media relations. Many Qwoted journalists now requests video responses as a service to confirm they’re speaking to a real person, he said.

Eftychiadou agrees with that approach. “I always tell my team, pick up the phone. Talk to people,” she said, noting that a critical part of FINN’s policy is never sending an AI-generated pitch or response directly to a reporter. That also helps them catch any content errors or hallucinations.

“It’s about using AI as a tool, not a crutch,” she said. “We can use AI to do the heavy lifting of analysis, which allows us to focus on what matters: building meaningful relationships with the media.”

Casey Weldon is a reporter for PR Daily. Follow him on LinkedIn.