Emerging Innovations in Journalism

I tested how well some AI tools actually work for journalism. Here's what I found, published today in the Columbia Journalism Review. 🤖🗞️ 👉 https://lnkd.in/eU4DVJsa As a reporter, I’ve often asked myself: Can I actually trust AI tools to support real journalism work? For me, and for folks like Hugging Face’s Florent Daudens, The Washington Post's Jeremy B. Merrill, and Sahan Journal's Cynthia Tu—“vibe checks” aren’t enough. I teamed up with a group of amazing researchers to run structured tests on some of the most popular AI tools. We used real-world editorial tasks, summarizing government meetings and reviewing scientific research, to see how these tools actually perform. The results? Surprising, frustrating, and occasionally impressive. 📝 Summarizing Local Government Meetings This is bread-and-butter work for many local journalists. Here's what we found: For short summaries (~200 words), tools like ChatGPT-4o, Claude Opus 4, and Perplexity Pro did surprisingly well, often capturing more facts (and hallucinating less) than the human-written summary we used for benchmarking. For longer summaries (~500 words), the quality dropped fast. On average, the tools retained only about 50% of the facts, hallucinated more, and missed key details. ChatGPT-4o had the most consistent and accurate output, with the lowest hallucination rate and best user experience. So: AI can help with quick recaps—if humans are verifying the work. But for more in-depth reporting, it still needs a human doing the work. 🔬 AI & Scientific Research: Not There Yet We also tested newer AI tools designed to help journalists and researchers make sense of academic work, especially tools that promise to surface related studies or verify the importance of a finding. Most tools surfaced less than 6% of the citations included in expert human literature reviews. Across the board, results were incomplete, or just plain wrong. 100% do not recommend (yet). Huge thanks to the brilliant team behind this work: Sophia Juco, Sandy Berrocal, Nneka Chile, Julia Kieserman, Jiayue Fan, Emilia Ruzicka, Mona Sloane, and Michael Morisy 🙌 (and anyone I may have missed!). I’m especially grateful for funding and support from the The Patrick J. McGovern Foundation, Vilas Dhar, and Nick Cain, who are deeply committed to journalism’s future. Next steps: If you’re experimenting with AI in your reporting—or you’ve read the piece and have thoughts—I’d love to hear from you. Drop a comment 👇 or shoot me a message. I'm also looking to connect with others interested in developing AI benchmarking standards for journalism, to help folks test tools more easily and responsibly. Burt Herman, Paul Cheung, Aimee, Nikita Roy, Silvia DalBen Furtado, Nicholas Diakopoulos, Jeremy Gilbert, and many others, I see you! #AIinJournalism #MediaTech #Journalism #AI SABEW Investigative Reporters and Editors Global Investigative Journalism Network Online News Association MuckRock Foundation, Tech Policy Press

Agentic AI journalism has arrived. According to the UK Press Gazette this morning, Mediahuis, one of Europe's largest news publishers with 25 titles across five countries, just revealed it's experimenting with a chain of AI journalism agents to produce routine "first-line" news. Not just one AI tool. A full agentic AI pipeline: commissioning, writing, legal checks, fact-checking, multimedia sourcing, and discourse monitoring - all handled by specialised AI agents before a human journalist reviews and publishes. The goal? Free their (currently) 2,000 human journalists to focus on "signature journalism" — investigations, interviews, community-connected depth reporting. What does this mean for PR and communications professionals? How long before: 1. Your press release may be triaged by AI first. Mediahuis is building curated source databases — wire agencies, parliaments, think tanks, political leaders on social. If your organisation isn't in those source pools in a structured, machine-readable way, you may not even make the first cut. Being findable by validated AI system sources may become as important as knowing the right journalist. 2. The two-tier newsroom needs a two-tier pitch strategy. Routine announcements will increasingly flow through AI-mediated workflows. But "signature journalism" — the pieces that build reputations and break stories — still requires human relationships. Know which tier your story belongs to, and invest your time accordingly. 3. AI monitoring is now part of the editorial cycle. Mediahuis's monitoring agent tracks public discourse around published stories. When polarisation spikes, it flags the topic for deeper editorial investigation. That means how audiences react to initial coverage can now algorithmically trigger follow-up journalism. The crisis response window just got shorter and more complicated (if that's possible). The multi-agent workflow Mediahuis describes - commissioning, producing, checking, monitoring - maps directly to how many PR teams operate. Is there an opportunity to apply similar thinking to comms content production: use AI for the routine, preserve human expertise for the strategic? Though fewer routine journalism roles will mean an even thinner pipeline of experienced reporters long-term. And if multiple publishers adopt similar AI systems drawing from the same source databases, do we risk even more homogenised news coverage? What happens when dealing with agentic AI journalism systems becomes the norm? What changes are you already seeing in how newsrooms handle incoming stories? As ever, welcome your comments below. Read the original Press Gazette article here: https://lnkd.in/eZ5_SgpS

🤖 This small Norwegian newsroom taught AI to think like their best reporters. Fascinating conversation with Lars Adrian Giske and Rune Ytreberg at iTromsø, a 25-person Norwegian newsroom that's cracked the code on using AI to strengthen local journalism. Their secret? Turn veteran journalists' gut instincts into code. Here's how: "We asked [our expert reporters] to evaluate a huge dataset of public documents, giving it a thumbs down or thumbs up. Is it newsworthy or is it not?, explains Lars Adrian. They have that expertise, it's become a gut feeling. They've built that judgment over time." The AI system DJINN learned from these decisions, picking up the subtle patterns that make a story worth pursuing. Now, when faced with 400+ daily municipal filings, DJINN surfaces exactly what these veteran reporters would have flagged as newsworthy. It's not just scanning documents – it's replicating decades of newsroom expertise. The impact? Summer interns (fresh out of j-school!) produced 5 front-page stories in their first week using DJINN. They were breaking major urban development stories. 3 key lessons for newsrooms looking to innovate with AI: - Start small, prove value fast: iTromsø began with a simple property bot tracking real estate transactions. It worked so well that 70 newspapers across their parent company adopted it. This early win gave them the credibility to pursue bigger AI projects. - Build domain-specific tools and get your experts to annotate data: Instead of chasing general AI solutions, iTromsø creates specialized assistants for specific beats (urban planning, healthcare, fisheries). Each tool embeds the expertise of senior reporters, making it easier for new journalists to develop beat expertise. - Let editorial lead tech, not vice versa Every AI tool at iTromsø starts with journalists' needs. Their development process begins with extensive workshops with beat reporters to understand their workflows and pain points. The tech follows the journalism, not the other way around. The best part? More time for actual reporting. Instead of spending hours manually scanning documents, journalists can focus on verification, investigation, and telling stories that matter to their community. This is how we future-proof local journalism. One smart tool at a time. 💡 Want the full story? Listen to the latest Newsroom Robots podcast episode with Nikita Roy.

#GenAI is reshaping journalism —but how are journalists actually using it, and how do audiences react? This report brings together a multi-country study and dives into AI-generated content in journalism, the perceptions of journalists and use of GenAI, and the perceptions of audiences.  ➡️ Key Findings - AI bias can take many forms - AI tools / models are almost always frustratingly opaque - More acceptable to use AI illustrations versus using AI as a replacement to camera-based journalism - Concerns identified about the potential of AI-generated content to mislead - Concerns identified about the effect that GenAI will have on society - Minority of the outlets interviewed had GenAI policies in place - Transparency about when and how AI is used is important - Only a minority of interviewees were confident they had encountered AI-generated content - Consumers with AI experience tend to be more comfortable with its use in journalism ➡️ Use Cases - Enriching and brainstorming - Editing - Creating ➡️ Legal / Ethical Issues - Mis/disinformation - Labor displacement - Copyright - Detection difficulties - Algorithmic bias and reputational risk Report Source: https://lnkd.in/eMEf6Gne If you're interested in GenAI in Journalism, I also recommend this report from Reuters Institute for the Study of Journalism https://lnkd.in/enSuNQUK

From the latest Adriana Lacy Consulting newsletter | The Washington Post announced a content partnership with OpenAI, making its journalism available within ChatGPT. Under this agreement, ChatGPT will surface Post content, including summaries, quotes, and links to original articles, when relevant to a user’s query. For The Post, it’s a new avenue for distribution. For OpenAI, it’s a chance to ground AI responses in credible journalism. On one hand, these partnerships promise broader reach. ChatGPT and other AI platforms engage hundreds of millions of users each week. Getting journalism in front of those users, in a context that includes links back to original articles, is potentially powerful for audience development and brand visibility. But there are real risks. Will readers click through to the full article, or stop at the summary? Can AI platforms accurately reflect the nuance of quality reporting? And what does it mean for journalism if distribution becomes intermediated by tools publishers don’t control? For journalism to survive and thrive in an AI-driven world, publishers must not only experiment but they must lead. I’ll say that again. They must lead! That means proactively shaping partnerships with technology companies, insisting on transparent attribution, and investing in their own AI infrastructure to retain editorial control. The Washington Post���OpenAI deal is a signal moment, not a one-off. It suggests that the future of journalism isn’t a fight against AI but perhaps a strategic push to ensure journalism’s values, standards, and economics guide the evolution of these tools.

One of the most significant shifts in journalism today is that top reporters are abandoning prestigious institutions to build something entirely different. ➡️ Oliver Darcy left CNN → built Status Media to 80,000 subscribers ➡️Jessica Yellin left CNN's White House beat → reached 1.5M followers with News Not Noise ➡️ Leigh Ann Caldwell left NBC/WaPo → joined Puck's specialized journalism model What struck me most wasn't their success outside traditional media—it was their clarity about why proximity beats mass broadcasting every time. These journalists are establishing new principles for trusted media: ✓ Direct accountability to audiences, not institutions ✓ Human curation over algorithmic amplification ✓ Community-level relationships over mass broadcasting ✓ Specialized expertise over generalist coverage Kevin Merida, former executive editor of the LA Times, emphasized a crucial piece: "Build our local news ecosystem. That's where we've seen the vanishing breed of credible news organizations." His perspective reinforces that trust is rebuilt from the ground up, community by community. The revelation? In an era of information warfare, the most subversive act might be abandoning the fortress to build trust zones instead. Go niche, go local, put your ear to the ground. For those building businesses or leading organizations: What can we learn from journalists who chose relationships over reach? How might "trust zones" apply to your industry? #Media #Trust #Journalism #FutureOfMedia #MilkenInstitute

What are the biggest tools, trends and behaviours set to shape journalism in 2026? I listened to the Reuters Institute for the Study of Journalism’s podcast on this earlier today. It unpacks their recently published report, and while AI and video are prominent themes, what interested me most was how newsroom leaders are now thinking more soberly about AI in practice. In the past couple of years, newsroom leaders and journos have become increasingly aware that AI (and video) can’t just be shiny add-ons – they’re forces that are actively reshaping what news looks like, how people consume it and how people find it. On AI, it looks like the mood is mixed. People are understandably worried that it will continue to decimate referral traffic, but they’re also noticing the potential benefits of implementing AI in newsroom workflows. Many are seeing how it can change the production of journalism – from faster investigations to less time spent on the invisible grunt work. It also raises questions about how we add value in areas like summarisation and explanation, where AI can handle the raw processing, but human editors are crucial in deciding what matters, what leads and how stories are framed. In my own work, that’s shown up most clearly in data analysis, where AI has helped me make better sense of trends and patterns in our coverage. Some fear the technology will overtake traditional, human-created journalism entirely. But as we enter a world of AI summaries and synthetic content, the value of original reporting, deep investigative work and human voice might actually increase. It’s something we’re already witnessing with the likes of local media company Mill Media. Joshi Herrmann and his team are doing brilliant work reviving local investigative reporting. At a time when many local or regional publications are disappearing into the ether, Mill Media’s titles are stepping up, unearthing scandals and growing a dedicated subscriber base. AI can certainly assist in that investigative work, but human endeavour very much sits at the heart of it. I’ll share my musings on video another time, because this post has ended up longer than I anticipated... The report is definitely worth a read if you’re interested in where media goes next, and the podcast, with Mitali Mukherjee and Nic Newman, summarises it neatly in 30 minutes. ❓ I’m interested in what my network – both those inside and outside of journalism – thinks about the use of AI in journalism. Will it ultimately damage or enhance reporting, and how should newsrooms be reacting? Report: https://lnkd.in/ghkvy-sj Podcast: https://lnkd.in/gA2Vk74U

As we look ahead to 2025, the landscape of journalism and technology is set for significant transformation. Here are key insights from the latest Journalism and Technology Trends and Predictions report by the Reuters Institute for the Study of Journalism and the University of Oxford: 1. AI Integration: 87% of newsrooms report being transformed by generative AI, focusing on back-end automation, personalization, and content creation with human oversight. 2. Shifting Platform Dynamics: Publishers are reevaluating their platform strategies. Sentiment towards X/Twitter has plummeted (-68 net score) while emerging platforms like Bluesky (+38 net score) are gaining traction. 3. Revenue Diversification: While subscriptions remain crucial, 36% of publishers expect significant income from licensing to tech/AI companies - double last year's figure. 4. Innovation Focus: News organizations are exploring new products, including games, educational content, and youth-focused offerings. 5. Talent Challenges: While 81% are confident in retaining editorial talent, there are concerns about attracting and retaining tech talent, especially data scientists and software engineers. 6. Audience Engagement: Combating news fatigue and adapting to conversational AI interfaces are key priorities. The rise of Bluesky in the social media landscape is notably interesting. This emerging platform has captured the attention of major news organisations, including The New York Times, The Guardian, and The Financial Times, which have already established a presence on Bluesky. These outlets frequently report higher engagement levels on Bluesky than on their accounts on X or Threads. This trend suggests a potential shift in how news is distributed and consumed on social media. Bluesky's growth key factors: 1. User migration: Dissatisfied users from other platforms, especially X/Twitter. 2. Robust moderation: The platform offers stronger content moderation policies. 3. Decentralized architecture: This feature gives users greater control and customization options. As we navigate these changes, the ability to balance technological innovation with core journalistic values will be crucial for the future of news media. This trend indicates a possible shift in how news is distributed and consumed on social media. What do you think about these insights? How do you perceive their impact on the way businesses communicate? Dinis Guarda Thomas Power Neil Milliken Sally Eaves Paidi O Reilly Tony Moroney Azita Esmaili Lynette Jackson 🚀 Ellen Schramke Ophelie Janus Dr. Ralph-Christian Ohr Kevin O'Donovan Paul Hobcraft João Morais Dalila Carvalho Marcus Borba Neville Gaunt 💡⚡️ Timothy "Tim" Hughes 提姆·休斯 L.ISP Efi Pylarinou Tamara McCleary Audrey DeSisto ARLENE NEWBIGGING GRADY Mei Lin Fung David Bray, PhD Frances West Eveline Ruehlin Jorge Cunha Luis Lancos Peter Torres Fremlin Shruti Pushkarna #Journalism #Media #Innovation #ArtificialInteligence

Reach PLC won’t exist in a few years. With the rise of LLMs and Google’s AI Mode & Overviews, people are getting the information they need without ever clicking through to the site that created it. No clicks means no ad revenue, and Reach PLC’s business is built entirely on that. Maybe that’s not a bad thing (hear me out). If AI gives you the answer instantly, there's no incentive to churn out filler content anymore. In the 90s, people paid for news, they bought newspapers, and signed up for 'email bulletins'. That model has been making a comeback. Take The Mill (manchestermill.co.uk) by Joshi Herrmann or former Welsh Affair's editor at Wales Online Will Hayward's newsletter (https://lnkd.in/d7RmxuNe), two of many examples. We’re coming full circle. Instead of chasing SEO traffic, we’re seeing a shift toward smaller, more focused, reader funded media. Substack, Patreon, and paid newsletters are thriving. People pay for quality, relevance, and personality. Reach PLC and its peers were built on volume, more articles, more clicks, more ads. But that model collapses when AI becomes the middleman, or the destination. In a world where AI surfaces the answer instantly, there’s only room for two types of content. 1) Truly original journalism worth paying for 2) Hyper-niche perspectives AI can’t replicate Churnalism thrived in the traffic economy, but that economy is ending. The internet broke journalism in some ways. AI might be what forces it to fix itself.