Keynote speech by Jean Philip De Tender, Deputy Director General/Director of Media, EBU, at the ABU-Rai Days conference in Naples, 11 June 2024.

Good morning, dear friends, colleagues, and anyone who’s here because they care about the future of public service media.

I first want to say a big thank you to Simona Martorelli and RAI’s international relations team for bringing us together for this important discussion today.
 
It’s regrettably rare that we get to meet up with colleagues from the Asia-Pacific region, and it is always a pleasure to meet wit the ABU’s esteemed Secretary General, Nadeem Ahmed and the team.

So I’m very pleased to have the chance to tell you about where the EBU stands in this exciting, inspiring — and sometimes worrying — conversation about AI.

Because the relationship between PSM and AI never stands still; as this powerful technology evolves — with leaps forward on an almost weekly basis now — so does the way we relate to it, learn about it, and integrate it into our professional lives.

The subject I was asked to cover today is ‘What is AI, where did it come from, and where is it going?’ And because I nearly always do as I’m told, I will look at all three questions — although to be quite honest, I think the question that really matters is the third one – where’s AI going?

Because the fact is, wherever it came from, AI is here, it’s not going away, and it’s already changing everything. And what we need to focus on is working out the direction it’s heading in.

If we can do that, we’ll give ourselves a fighting chance of taking control of our future as public service media. Because AI is fundamentally rewriting how people create, consume, and understand information. And PSM — with the core mission of informing, educating, and entertaining the public — is at an inflection point in that change.

Is AI a Trojan Horse, eroding trust and objectivity? Or can it be harnessed as a tool to strengthen PSM's vital role in a complex and shifting media landscape?

The good news, ladies and gentlemen, is that the answer to this question lies in our hands.
 
So, where did it come from?

The early days of AI in PSM were driven by the search for efficiency. Algorithms streamlined content management, automating tasks like news aggregation, powering recommendation engines. This freed up precious human resources for more complex journalistic work. But before long, PSM noticed the deeper storytelling potential of AI.

Natural Language Processing tools helped analyse vast datasets, uncovering hidden patterns and stories within them.
Speech recognition software facilitated content accessibility, ensuring inclusivity for all audiences.

Today, AI is a multifaceted reality touching many areas of the PSM space, including:

News
AI tailors news feeds to individual tastes, ensuring audiences get personally relevant information. 

Great – but the danger here is filter bubbles and echo chambers — we must strike the balance between personalization on the one hand, and exposure to diverse viewpoints on the other. 


Content Creation
AI is handling captioning, editing, and transcription, allowing broadcasters to create content faster and more efficiently. But there are ethical risks around potential biases in algorithms, and the need for human oversight to ensure journalistic accuracy and integrity.


Fact-Checking and Verification
PSM are using AI to analyse masses of information to spot potential misinformation and fake news. Yet total dependence on algorithms for fact-checking is not as reliable as the crucial, nuanced skills of human journalists when it comes to investigative reporting.


Audience engagement
Machine learning algorithms can study viewer behaviour and enable PSM to tailor content and target specific demographics. Yes, this fosters deeper engagement — but it also raises questions about potential manipulation and the so-called ‘commodification’ of audiences.


And finally: Archiving and Retrieval
AI-powered search engines can sift through immense archives, giving fingertip access to historical content. But, again, the potential for algorithmic bias means we will always need qualified human curation and editorial oversight.

Now, let me tell you about the EBU and AI.

When the EBU set out its strategic priorities for 2024, the first among them was to ‘Provide thought leadership in the exploration and optimization of AI strategies and tools’.

In other words, artificial intelligence is the number one issue for us, and for many of our members.

So, here’s what we’re doing about it:

Firstly, we’re helping our Members shape smarter strategic approaches to AI, starting from the baseline that this technology is — to put it bluntly — do or die.

Let me ask you a question. In the very early days of social media, how many of you here reacted sceptically, or at least put yourself in the wait-and-see camp?   

With AI, we simply don’t have that luxury. Because artificial intelligence is radically changing our processes and workflows, and anyone who gets left behind will fade into irrelevance.
 
That’s why we’ve developed a PSM AI Strategies paper by talking to 15 Members that are already quite far along the AI road.

Members can download it from our website, and it’s a useful, living document that’s helping other members benefit from the learning experiences of those 15 early adopters.  

Now, with so much at stake, AI can be a divisive topic. Will it be a job creator, or a job eradicator? There are big unknowns around ethics and trust, as well as policy and legislation. How will public service media work with the tech giants who, by cutting corners to get ahead in the AI race, are jeopardizing the trust between PSM and their audiences? 
 
At the EBU, we are mindful of these hazards and we’re working to address them.
 
But, on balance, we lean into optimism, because of the enormous opportunities this frontier technology promises for efficiency, creativity, and bringing PSM closer to audiences. 
 
Today, like many of our Members, we’ve triggered numerous wide-ranging workstreams to unlock the benefits of AI. We’re building the EBU as a centre of AI excellence, expertise, and learning for public service media. 
 
-       Our first AI Summit, in December 2023, brought together 400 leaders and AI practitioners to discuss the opportunities for collaboration and learning. It was so impactful that we’re making it annual, and the next one’s on 6th December, in Lausanne, Switzerland. If you can come, please do.
 
-       On the regulatory front, we’ve mobilized our Brussels team to keep PSM across the EU’s legislative portfolio, and legal changes that will affect our members.
 
-       In terms of learning and upskilling — another key focus area — we’ve launched the EBU Academy School of AI, which is training Members in new specialisms, from using generative Al to create content, to prompt design, ethics, and audience personalization. 
 
-       And we have an AI Ethics Group, which meets monthly to discuss the ethical implications of PSM using AI, to share guidelines and principles, develop practical tools, and establish coherent EBU positions. 



Because ethical use of AI is a non-negotiable for PSM, and here’s why:

1. Transparency: 
Audiences have a right to know how AI is used in content creation and distribution, so PSM must be clear about the role of algorithms and human intervention.

2. Bias Mitigation: 
Data sets used to train AI need to be as diverse and representative as all PSM output. Above all, they must avoid perpetuating harmful biases.

3. Algorithmic Explainability: 
Understanding how algorithms arrive at their decisions is crucial for fairness and accountability.

4. Human Oversight: 
Human judgment remains essential for ensuring the quality, accuracy, and ethical integrity of content.


Together, we’re working as a community to define a compelling common direction on AI, to foster Members’ creative AI developments, and develop a repository of ideas on AI for content creation and audience engagement. 
 
Ladies and gentlemen, AI represents change, and change can be scary. But if we embrace AI and harness its power cleverly, transparently, and responsibly, it will be to our collective advantage and, crucially, to the benefit of the public that we serve.

So what does the future hold?

Well, looking ahead, I see only one possible scenario: AI will become even more deeply enmeshed in the fabric of PSM – their workflows, their creativity, their content output, and distribution.

We’ll see more AI-powered investigative journalism, where AI analyses complex datasets, uncovers hidden patterns and spots elusive leads for investigative journalists to chase.

We’ll see more multilingual creation, with AI translating content in real-time and making PSM programmes accessible to an unimaginably wide global audience.

And — why not? — there’s every chance we’ll see more personalized documentaries and storytelling, where AI customizes content to individual viewers based on their interests and viewing histories, creating a more immersive and engaging experience.

The possibilities are staggering. But the AI conversation is one that always goes ‘on the one hand, A…and on the other hand, B’.

Because none of its benefits come without risk.

Navigating our AI future will require anticipating and overcoming many challenges around issues like:

Regulation and Governance: It will be crucial to develop ethical frameworks for AI in public service media. And I would stress that PSM need to take a lead role in shaping these frameworks.


Job displacement: The uncomfortable reality is that AI-driven automation may lead to job losses in the media industry. PSM must invest in retraining and upskilling programmes, like those offered by the EBU School of AI, to equip their workforce with the skills needed for an AI-driven media world.


And of course, the misinformation war.
As deepfakes and other forms of synthetic media become more sophisticated and more dangerous, we must fight fire with fire, and use AI to detect and combat misinformation.

I said before that the EBU leans into optimism, and I meant it.

So, to wrap up, I’ll leave you with a few optimistic thoughts.

The future of public service media in the AI age is not one of machines replacing journalists and creatives.

It's about forging a symbiotic relationship where AI empowers PSM to fulfil their mission more effectively and efficiently.

By embracing AI responsibly, transparently, and ethically, PSM can ensure that the ‘public square’— both physical and digital — remains a vibrant, representative space for informed discourse, diverse perspectives, and a deeper, participative understanding of the world around us.

Thank you.