Generative AI in the Newsroom: Friend or Foe?: A Panel Discussion Hosted by FleishmanHillard
FleishmanHillard recently hosted a timely and deeply illuminating panel discussion on the impact of Generative AI (GenAI) in modern journalism. Held at the Axel Springer Haus in downtown New York City, the event brought together senior newsroom leadership from leading publications, including AdWeek, Axios, Business Insider, Harvard Business Review and POLITICO. Moderated by Ephraim Cohen, FleishmanHillard’s global managing director of Media, Platforms and Storytelling, the panel addressed the challenges, fears and opportunities GenAI presents for newsrooms worldwide. The event attracted an audience of more than 40 attendees, including senior communications professionals from top-tier companies and clients like Novo Nordisk, Meta and Samsung, among others.
To accompany the panel discussion, FleishmanHillard also published a study on the current state of GenAI in the newsroom – from the early adopters and the critics to the initial successes as well as the emerging cautionary tales.
Panel Highlights
The discussion featured a stellar lineup of panelists who provided deep insights into GenAI’s evolving role in journalism including:
Aja emphasized the importance of distinguishing between commoditized content and unique reporting delivered by expert journalists. She highlighted how GenAI can streamline workflows but stressed the necessity of maintaining trust and community:
“Trust is at the center of what we do. We are reorienting our newsroom and business around subject matter experts that can deliver something specific to them, their experience, and their skillsets. If we orient around trust and expertise, we are confident we can weather the technological storm that is coming.”
Cadie shared how Business Insider is leveraging GenAI to enhance productivity, such as using AI to digest meeting notes or news reports to decide on story depth. She also highlighted AI’s role in transcribing podcasts and shows, freeing journalists to focus on more critical tasks. Cadie’s approach reflects a balance between innovation and journalistic integrity:
“No one likes change, and the integration of Generative AI is a big change. One of the blessings of working with journalists is that they are naturally skeptical of everything, so we trust them to be careful with new tools. As the executive editor, I try to show them the possibilities of how gen AI tools can make their lives easier and allow them to focus on the things they want to be doing.”
Ryan highlighted the need for journalists to understand the craft of journalism before integrating AI tools into their work. He cautioned against giving AI tools to junior reporters without proper training. Ryan also pointed out practical uses of AI in generating drafts for complex advertising content. His thoughts reflected a careful and measured approach to AI adoption:
“Retaining readers’ trust is fully in our control, regardless of Generative AI. AI used to be a dirty word in the ad industry. But now, even if we tried, we couldn’t not cover it.”
Francesca focused on the importance of transparency in AI integration. She discussed how Politico is using AI to do things like generate bill summaries and enhance storytelling while maintaining transparency with their readers. Francesca’s emphasis was on building and maintaining trust through open communication:
“When we think about building trust and protecting against erosion of that trust and accuracy, we have to be self-aware that we are not the single voice. The integration of Generative AI into our storytelling strategies brings about questions. How did we get the reporting? Who was involved? So the more we can focus on transparency, the better.”
Adi shared Harvard Business Review’s investment in developing a proprietary AI model to ensure accuracy and reliability. He highlighted the duality of AI as both a powerful tool and a source of concern. Adi’s insights captured the cautious optimism surrounding AI’s future in journalism:
“Part of the reason we’re investing in generative AI internally is because it’s going to be table-stakes. Another part is because existing LLMs (Large Language Models) are error prone. If we can build a model that doesn’t hallucinate and is based on our own content, then we have a clear alternative to big tech platforms. But obviously, trust is everything.”
Key Learnings
- Trust and Expertise Are Crucial: Emphasize the irreplaceable value of experienced journalists to maintain reader confidence amid GenAI integration.
- Transparency Enhances Credibility: Ensure clear communication with readers about GenAI’s role in reporting processes to build and maintain trust.
- Balanced Integration of AI: Utilize GenAI to enhance productivity while maintaining human oversight to ensure accuracy and quality.
- Guardrails and Ethical Considerations: Implement strict measures to prevent GenAI errors and prioritize ethical considerations, including privacy and misinformation.
- AI as a Tool, not a Replacement: View GenAI as a supportive tool that enhances, rather than replaces, the core human elements of journalism.
