BBC’s AI Strategy, Affiliate's Comeback, Hearst’s Local News Bot, and more...
A shout out to today’s sponsor, Writers’ Bloc — Publishers are losing millions as their work gets scraped, stolen, and resold. Using digital fingerprints, Writers’ Bloc turns every article into a protected, monetisable asset on the blockchain. It protects your IP, tracks provenance, and unlocks new licensing revenue automatically.
For smaller publishers, implementing AI can be quick, transformative and friction-free. The same can’t be said for publishers with many hundreds of staff. Where do you start? And what are the guardrails? This week’s Long Read looks at how the BBC News is implementing AI across its 5,500 staff, a herculean task by any measure.
Let’s jump in…
Judge Says AI Lawsuit Against Cohere Can Proceed
A who’s who of major publishers - Forbes, Guardian, McClatchy, LA Times, Politico, Vox, etc - is suing Cohere, a Canadian AI company that builds LLMs for businesses, for copyright infringement. Why Cohere? It’s a smaller target, more easily proven, and once a precedent is set, the publishers can take on the bigger AI platforms. #strategicmove #NewsMediaAlliance
Hearst’s New AI Tool to Expand Local Coverage of Public Meetings
Hearst Newspapers has developed ‘Assembly’, an AI tool that automates the reporting of public meetings. It’s damned smart: 200 custom Web scrapers scan government websites for new meetings, spotting and transcribing the proceedings. Reporters also receive email alerts when a predefined keyword is detected. The tech even enables reporters to have “conversations” with the meeting transcript itself…
Is Google About To Go Full AI Mode?
AI Mode is not the same as AI Overviews, and is a far more advanced, interactive search experience where people spend 3x longer. Google’s product lead has suggested that it will “soon” become the default search option. The ramifications are profound, with publishers needing a strong presence across Reddit, Quora, YouTube, etc, to underpin themselves as a ‘trusted source’. Ominous: “Paid visibility in AI results will grow rapidly.”
Forbes AI-Powered Paywall Shows Early Promise
Witnessing a 40% drop in traffic from search, Forbes has launched an AI-powered paywall which learns from reader behavior and adjusts subscription offers in real time. In just a month since launch, it’s seen mobile conversions increase 400%, desktop conversions jump 174%, and acquisition costs decrease by 13%. Key quote: “The goal is to offer the right product at the right time.”
Daily Mail Says Google AI Overviews Have Killed Click-Throughs
When an AI Overview is triggered, click-throughs from that specific search keyword are DOA for the Daily Mail. How bad? 80-90% lower when an AI Overview appears, even when the Mail ranks on the first page of Google for that same keyword. It’s a steep drop from the 56% lower CTR it disclosed in May. P.S. Bloomberg CEO says its going to zero…prepare accordingly.
Fox News Builds Digital Twin to Experiment with AI
To test, experiment and build AI tools, Fox News has hired Palantir to create an entire digital replica of its business. The model mirrors the workflows, data, tools and systems the company uses in everyday reporting, allowing in-depth AI experimentation in a ‘live’ newsroom setting. Note: New research shows readers want human involvement, guardrails, transparency if AI is used in news reports.
LinkedIn: From Ensh*ttification to Algorithmic Redemption?
LinkedIn’s been drowning in shallow, vacuous crap. We’ve all done it: “I don’t normally post these…”, alongside a photo of a dog in a bandage. With the average user having 1,300 connections, that’s three shaggy dog stories each day. LinkedIn has now changed its algorithms to focus on business acumen and real advice, prioritising authority over activity. For B2B publishers, it’s a fresh opportunity...
The Atlantic Shifts Its Focus: Less Traffic, More Intentionality
Chief Growth Officer Megha Garibaldi says the days of “tens of millions stumbling upon us” are over and it’s now all direct engagement: sharper newsletters, follow-your-favourite-journalist tools, smarter paywalls and, *sigh*, even games. With 1.4M subscribers, retention is the priority built on ‘human-centered relationships’. The Atlantic’s reporting has also improved beyond measure…
Why Affiliate Marketing Suddenly Matters Again For Publishers
AI platforms (ChatGPT, Perplexity et al) are changing how people find products online, but this time brands can’t kick publishers aside using PPC (paid ads). It means affiliate marketing - researching & writing useful content, creating product recommendations - is back. With 84% of online publishers still using affiliate links, it’s time to dust down your skill set…
Five Critical Trends Reshaping Marketing
John Rahim is one of those rare individuals who spans both media and marketing, and his teardown of WARC’s Marketing Playbook 2026 is a must-read. #1 trend? The term “middle-class” is now meaningless, it’s disappeared, with the U.S.’s top 10% accounting for half of all consumer spending. In short, go premium or go value, but pick a lane…the middle is dead.
AI Tool of the Week: Flourish
Design stunning interactive charts, maps, and even animated stories using research data — no code required. Quick to create, this is an unbeatable tool for publishers with limited resources or design capabilities. Part of Canva, and still in beta, it points to where visual tech is heading.
Webinar Series: Nearview Media High Value Publishing
Eric Shanfelt is one of the U.S.’s leading publishing experts helping media companies build revenue and audiences across B2B, consumer, and regional publishing. Packed with operational insights based on his ‘in-the-field’ experience, the webinars are free. Each Monday | 8pm UK, 3pm EST | Sign up.
Event: INMA World Congress of News Media
Titled “From walls to bridges: news media reimagined”, INMA’s 2026 Congress combines study tours, seminars, a two-day conference, the Global Media Awards, and private AMA sessions with INMA subject matter experts. If you’ve any energy left, Berlin also offers its delights. May 4-8 | Berlin
Inside the BBC’s AI Strategy: Practical Lessons for Publishers
Let’s put aside recent controversies – there’s always one at the Beeb – yet it doesn’t hide the fact that it remains one of the largest media organisations in the world, with over 5,500 people working in its news division alone. Its reach? 426M people access BBC news, worldwide.
With the implementation of AI now a strategic necessity, where does an operation of the size of the BBC begin? What problems do you solve first? And why? And how do you ensure new tools support, rather than erode, editorial standards? *cough*
These are the questions facing Olle Zachrison, Head of News AI at BBC News. Speaking last week on Newsroom Robots – a podcast hosted by Nikita Roy at Harvard Innovation Labs – he describes the fierce challenge of helping a sprawling, 24-hour global newsroom adopt AI without losing focus or trust.
Zachrison leads a relatively small, dedicated team tasked with embedding practical, ethical applications of AI into daily workflows. Its mission is not to replace journalism but to support it with useful, editorially sound tools.
“We’re an acceleration team for the responsible use of AI.” — Olle Zachrison, Head of News AI, BBC News
A framework with four clear themes
Zachrison explained that BBC News organises its AI work under four themes: boosting productivity, reformatting content, augmenting journalism, and improving user experience. This gives structure to the broadcaster’s experiments and allows teams to work toward shared outcomes.
Boosting productivity includes using AI for transcription, tagging and translation. These tools are designed to cut down repetitive tasks and help journalists spend time on higher value work. Translation is one of the more advanced projects, with BBC World Service working across 42 languages.
Reformatting content is about adapting what already exists for different platforms and audiences. That could mean converting articles into short podcasts, resizing videos for different social media channels, or creating live text pages from audio feeds.
One example is My Club Daily, a sports podcast that uses AI to generate short, club-specific audio updates from BBC Sport articles. Scripts are reviewed by editors and read using synthetic voices. The idea is not to replace journalism per se, but to deliver existing work in new formats that meet audience habits.
“We start with the article. That’s the raw material. Then we rework it for video, social or audio. AI helps connect that process.”
— Olle Zachrison
Building confidence and capability
When it comes to augmenting journalism, the focus is on tools that support strong original reporting. That includes sifting through large datasets, finding story angles in spreadsheets, or helping reporters test ideas faster.
Training is a core part of this effort. Staff use tools like Microsoft Copilot, ChatGPT and custom GPTs. Adoption is growing, but Zachrison believes peer-led learning has the most lasting impact. “It’s not enough if someone like me says this is useful,” he said. “It matters more when a colleague shares how it helped with a real story.”
He also pushed back on the idea that journalists are resistant to AI. Most are under time pressure and don’t get space to explore new tools. But when they do, the ideas often come thick and fast.
“A lot of the best ideas are already in the newsroom. We just need better ways to capture and act on them.”
— Olle Zachrison
Innovation that fits the brand
The final theme is user experience. Zachrison sees potential in using synthetic voice to improve access to news, especially for people who prefer audio formats. The BBC is also looking at conversational interfaces, where users can ask for summaries or updates in a more interactive way.
These projects are still in early stages, and Zachrison is careful not to overstate their impact. He cautions against treating innovation as a race to match what big tech firms are doing.
“People don’t come to the BBC for an AI experience. They come for credible news. That’s what we need to deliver, with the help of AI where it makes sense.”
— Olle Zachrison
The BBC’s approach contrasts with Fox News, which has taken a far more aggressive path. It’s hired Palantir to build a digital replica of its newsroom and embedded engineers directly into its editorial workflow. The aim is speed: build custom tools, streamline production, and protect IP through commercial-first contracts. It’s AI transformation at pace.
The BBC, by contrast, is playing a longer game, focused on structure, internal alignment, and trust. This slower approach helps keep editorial values at the centre and avoids the trap of launching products for the sake of novelty.
“Just because it’s an idea doesn’t mean it’s a good idea. Sometimes you need to kill an idea in its sleep.”
— Olle Zachrison
Key takeaways for publishers
Whether you’re part of a public broadcaster or a small regional outlet, the BBC’s approach offers clear lessons:
Work to a structure. Themes give shape to AI efforts and prevent scattered pilots.
Start from what you already have. Use AI to repackage existing content into new formats before building new products.
Balance top-down and bottom-up. Encourage innovation from editorial teams as well as product or tech leads.
Focus on adoption, not just tools. Training helps, but trust builds faster when colleagues share what worked for them.
Protect the core offer. Use AI to support credibility, not risk it. Human oversight remains essential.
The BBC’s model will not apply to every publisher, but its method offers a useful framework. The key takeaway? Focus on clarity, internal alignment and results you can measure. If a project does not work, drop it. If it does, build on it.
As Zachrison concludes, the real value is not in having the latest tools, but in knowing what to do with them.
🎧 Listen to the full podcast here 🎧
Hat tip: Sir Anthony Hopkins’ memoir, “We did OK, Kid” - A Wales lost in time but one I remember all too well #PontypoolFrontRow #Steelworks