A few months back, (Gen AI) technology and talked about some Gen AI test pilots that we would be working on. I want to update you on these pilots, talk about what鈥檚 next 鈥� and share some of our thinking in this area. All of this is guided by our three AI principles to:
- always act in the best interests of the public
- always prioritise talent and creativity
- always be open and transparent with audiences when we use AI to support content-making.
How have the Gen AI test pilots gone?
Over the last few months, we have conducted some test pilots to explore different areas where Gen AI could bring significant benefits for audiences and for staff. Most have been focused on content production and have been assessing how Gen AI can simplify production processes and reformat existing content 鈥� so more of our audience can enjoy it. Some were live trials 鈥� seen by our audiences - and others were behind the scenes.
On 大象传媒 Sounds, we鈥檝e been using AI tools to including In Touch, Access All, Profile, Sporting Witness and Economics with Subtitles. This is working well 鈥� the accuracy is impressive - and this trial has now been extended to include more programmes, including The Archers and The Today podcast. In parallel, we鈥檝e been exploring how AI can generate short-form animated sequences (instead of just static images) when showcasing programmes on 大象传媒 Sounds.
In Sport, we鈥檝e been looking at how AI tools can add more value online. This was focused on our coverage of the English Football League coverage and the hundreds of live commentaries provided on 大象传媒 Local Radio. Using AI tools, we were able to quickly generate transcripts of these commentaries 鈥� and identify the key moments of the match (the goal or the red card, for example). After being checked by our journalists, they were published as live text commentaries on the 大象传媒 Sport app. This has been popular with fans 鈥� so we will be looking to do more of this with other matches.
In news, we鈥檝e found Gen AI is most helpful behind the scenes in accelerating production processes. For example, we鈥檙e now using AI tools to provide much faster translations of reporting across our Language Services - again with the usual human checks. For example, we used AI to assist the translation of the 大象传媒 Mundo article on the 20th Anniversary of the 2004 Boxing Day tsunami for our 大象传媒 Brasil service. Using Gen AI means we can publish stories more quickly across the world
What have we learned so far? The Gen AI tools are impressive, the accuracy levels are improving and the pilots suggests we could unlock significantly more value for audiences. In the majority of cases, it鈥檚 also clear AI can assist us to do things more quickly but it does not remove the need for human oversight given the risk of error.
Being clear with our audience
We also know that AI-generated content can blur the line between the real and the artificial. As our research tells us, our audiences want to understand how AI is used in our work and the impact it has in the content that they watch read, watch or listen to from the 大象传媒.
Transparency is key to maintaining trust, and we are committed to being open about where and how AI is used. That鈥檚 why we鈥檝e introduced new guidelines for staff and external producers and suppliers for disclosing its use, informed by what our audiences have told us matters.
In all cases, we now label AI use in our content where there鈥檚 any risk that our audiences might feel misled 鈥� such as the use of AI 鈥渇ace-swapping鈥� techniques to anonymise contributors or the live text commentaries that I mentioned earlier. In news, we also disclose 鈥榖ehind the scenes鈥� use of AI where it has a significant impact on our content 鈥� such as when we use it translate articles.
Training our teams
We鈥檙e also training teams to use AI tools like Microsoft Copilot, Adobe Firefly and GitHub Copilot 鈥� to make our work quicker and easier, all guided by our AI principles.
Our R&D teams are continuing to invest in work in this area, including developing our own Large Language Models 鈥� tools that are specially trained on the 大象传媒鈥檚 own content and data so that any future outputs reflect our values and tone.
The impact of AI on the wider media
More broadly, the impact of AI on the creative and news sectors is a key topic that will be much debated in 2025. The 大象传媒 is a leading voice in these discussions, and we are championing and supporting the UK鈥檚 creative industries and the country鈥檚 news industry.
As you may have seen, the UK Government is currently consulting on changes to copyright in relation to the training of AI models. We will be responding in due course and we do not believe that 鈥榮craping鈥� of 大象传媒 data without out permission to train Gen AI models is in the public interest.
In parallel, we will continue to engage with AI platforms across the world to ensure the 大象传媒 鈥� and its brand 鈥� is not misrepresented in AI outputs. Wherever the 大象传媒 brand is displayed, it is essential that audiences globally can trust the content that is being provided.
What鈥檚 next?
This year, our work will accelerate 鈥� so expect to see further deployment of AI tools across the organisation. We will be fast-tracking a number of our test pilots and rolling them out more widely where we believe it is safe and valuable to do so. We look forward to sharing more with you in the coming months.
Thanks for reading.