At the Media, Production and Technology Show (MPTS) 2025, Erika Young (Reuters News Agency), David Candler (Reuters Imagen), and Dr. Daniel Chávez Heras (King’s College London) came together to explore the growing role of AI in reshaping media workflows. This handout captures the key insights and takeaways from that conversation, covering the opportunities AI presents and the challenges media organisations must navigate to use it responsibly.
Whether you’re managing historical footage or planning next-gen content delivery, AI is fast becoming an essential part of the toolkit.
AI underpins everything from tagging and indexing to distribution and monetisation.
For large archives, especially in sport and broadcast, it’s no longer optional. It’s critical.
Tasks like transcription, translation, synopsis creation and contextual metadata enrichment are now easily handled by AI, freeing up human effort for higher-level creative decisions.
David Candler: “Without AI, you cannot find the content you need. If you’re in a sport, you need to find it very, very quickly. And if you can’t find it, you can’t activate it, you can’t monetise it… We are investing $200 million a year in AI R&D and that’s trickling down to all parts of Reuters and Reuters Imagen.”
Erika Young: “Reuters has a long history in investing in the latest, cutting-edge technology to ensure we are delivering best-in-class content. Having a thorough understanding of the end-to-end content development cycle for Newsrooms, Sports organisations, and Content Creators, we have identified certain capabilities, when enabled by AI, unlock efficiencies and enable Clients to do more with both our and their content.”
AI literacy, both technical and cultural, is critical. Upskilling the workforce needs to go hand-in-hand with adoption.
Dr. Daniel Chávez Heras: “The educational part and the training part [is] almost as
important as the technology itself. How to upskill workforces, and who pays for that upskilling, [are] as critical as… having access to the latest computational models.”David Candler: “A lot of people are now embracing AI, as opposed to being very worried about it taking over jobs. I think [we understand] that AI is an amazing tool. At Thomson Reuters, we are given suites of AI tools to use. We’re actively encouraging it. It’s part of our personal KPIs, that we all get to a level of understanding of the use of tools.”
Erika Young: “Reuters benefits greatly from being part of Thomson Reuters (TR), whose technology teams, including TR Labs, are at the cutting edge of developing AI solutions for its clients. TR has created a proprietary platform where we are all encouraged to experiment with gen AI and to safely and ethically use AI in day-to-day work. Furthermore, Reuters benefits from TR’s commitment to invest $200 million a year in AI.”
AI search tools now go beyond keywords. Vector search, in particular, is powerful for identifying patterns across complex media sets, returning contextual results so a photo or moment in a video can be found in seconds.
David Candler: “The thing about search is there are two main camps… semantic search and vector search… Semantic search is using NLP (Natural Language Processing) to understand what you’re asking for, and it’s trying to deliver those moments. But vector search is using a mathematical algorithm to create vectors, and then it’s trying to identify similar vectors. We find that a lot more powerful [and] that’s where we’re going with Reuters Imagen… Good logging creates a good dataset, creates a good search.”
Most AI systems have been trained on recent media, limiting their accuracy with content that predates the last 20 years.
Dr. Daniel Chávez Heras: “The larger portion of their material is historical… Many AI products and services have been trained in the past decade or in the past 20 years… So, there is that challenge of [a] mismatch with the technology and with historical material.”
Monetisation is important, but so is cultural preservation, identity, and narrative control. AI models must respect this wider context.
Dr. Daniel Chávez Heras: “There are different kinds of value… there is the potential financial value that you can derive from having these assets. There is also the historical value, the cultural value… Some of that has been captured by AI models, but some of it has not.”
AI is not just another feature. It’s a foundational technology, one that is comparable to electricity or steel in its potential to reshape industries.
Erika Young: “Reuters has been using AI and ML (Machine Learning) for over a decade and was a pioneer when we started using automation to extract economic and financial data from websites and press releases (EAP, PLX). GenAI is the next phase of our history of using AI tools and we regard it as a breakthrough that offers potential to enhance our journalism and empower our journalists.”
Dr. Daniel Chávez Heras: “AI has some of the hallmarks of different technologies that have had this transformational capacity - electricity, the creation of plastic, the creation of steel, for example. But it isn’t enough to be technologically oriented or technologically savvy. You also need a broader understanding of the cultural aspects of these technologies… the way they impact, or they can potentially impact, and many of the unintended consequences that they may have.”
AI will impact traditional roles, especially entrylevel ones. Preparing for that shift means leaning into creativity, not resisting change. Rather than replacing humans, AI should augment expertise. The ultimate goal should be a collaborative model where control, ethics, and creativity are retained.
Dr. Daniel Chávez Heras: “Every revolution has casualties… and that may include job casualties, and there is no way around that.”David Candler: “A lot of our customers are in sports…they need content very quickly, and they need particular moments. So… auto highlight generation is a big topic - ‘Give me all goals by Saka in the second half of this game’ - then it will automatically find those and put it into a Collection so editors can do their thing.”
David Candler: “[AI] is here. It’s not going anywhere. So, it’s about being ‘on the loop’ and having control of these tools to help you do what you do. So, I think we need to embrace, manage, and focus on that creativity…We have to make sure that AI doesn’t deceive or mislead, especially in the news world.
Reuters Imagen uses AI to support rights holders and broadcasters with fast, automated content delivery, particularly in high-pressure environments like live sports.
David Candler: “A lot of our customers are in sports…they need content very quickly, and they need particular moments. So… auto highlight generation is a big topic - ‘Give me all goals by Saka in the second half of this game’ - then it will automatically find those and put it into a Collection so editors can do their thing.”
As you reflect on the session, consider where your organisation stands across search, workflow automation, training, and AI governance. Adopt tools that streamline your production, but always keep humans at the centre of decision-making. AI is here to stay, but its value will depend on how intentionally, and creatively, we use it.
The Reuters AI Suite is a gateway to smarter, faster, and more scalable media workflows. Designed to enhance everything from content discovery to distribution, the suite brings together proprietary AI tools and best-in-class third-party technologies. Whether you’re looking to accelerate highlight generation, improve metadata accuracy, or streamline translation and transcription, Reuters AI Suite empowers you to work more efficiently and creatively, without compromising on trust or editorial standards.
Producing great content is an increasingly cost and labour-intensive process, but that’s not where the work ends. For those producing broadcast media, their partners are demanding speed, they need to find an exact moment in video, so they get to market first with their fresh take on content.
At Imagen, we are going beyond traditional models of AI-generated metadata such as transcription to truly surface the meaning behind video. Find moments instantly with Imagen’s AI-driven content library.