News
On 12 November the media industry met in Hamburg for the AI Media Leaders Conference 2026, and Christoph was there for us. The day covered Europe's place between the US and China, brands that can no longer buy attention, and broadcasters looking to make money with AI. His notes.
Christoph Föhrdes, Managing Director
Hamburg's media senator Carsten Brosda opened the day with an assessment that didn't quite fit the usual pessimism: Europe doesn't have to just rebuild what comes out of the US and China. In agentic AI, for example, there is still room at the top, if money and people are invested now.
How wide the gap is in practice came through in the discussion between Stephan Scheuer (Handelsblatt) and Moonie Zhu (eTOC) on the AI models of the US and China. Adoption in China is far broader: companies test a lot and get it to end customers quickly.
Marco-Alexander Breit of Germany's Federal Ministry for Digital Transformation and Government Modernisation struck a different note in the panel on Europe's path to AI sovereignty. Germany's treasure, he said, is its industry, the manufacturing it is strong in. That data and know-how should not be opened up to US models; instead the country should find ways to use AI on its own terms.
In the panel “Between algorithm and brand”, Kristian Meinken (pilot) said brands have to set themselves apart clearly again and offer something you only get from them. Building a brand is getting expensive again, because attention can no longer simply be bought in a search engine when more and more answers come from an AI.
Jennifer Lachman reported from DER SPIEGEL in the same panel. The publisher has made content available over MCP as well, the interface AI assistants use to reach outside data; the first collection, she said, was material on the Nazi era. The hard part is the line between them: what is carefully researched SPIEGEL content, and what did the AI add?
“Beyond cost savings” was the title of the panel with Maximilian Orgonyi (RTL) and Kathrin Franssen (Intermate). RTL showed two examples. Bundesliga scenes are re-enacted in real time with Playmobil figures, in cooperation with Playmobil, and the audience gets a say, for instance whether a pirate ship should cross the pitch. It is, of course, also new advertising space. The second: historical archive footage from courtroom shows is reworked with AI into new cases.
Intermate uses AI for brand safety. Influencers' channels are checked continuously, for antisemitism, nudity and more, so a brand can react quickly to a shitstorm or see the risks when choosing influencers in the first place. At its core it is content tagging.
In the afternoon Fabian Massah (Filmfest Hamburg) moderated a roundtable with Yoko Higuchi-Zitzmann. She founded Dreamkite Studios, a film production studio built on AI from day one, and shared what she has learned and where the business case for AI film lies.
Many of these topics are familiar from our own projects: opening content to AI assistants over MCP without blurring the line between checked content and generated additions, and using AI so that your own data stays in the house. The full programme is available from nextMedia.Hamburg.
On 26 and 27 November the Symfony community meets in Warsaw, and five of us are going: Vira, Liubov, Björn, Dmitry Dubinets and David. Three tracks run in parallel, so we split up. Here are the talks we've picked and why.
Mistral released Large 4 as a public preview today. It is a 1.05-trillion-parameter Mixture-of-Experts model with 49B active parameters, text and vision input and a one-million-token context window. Mistral trained it from scratch on 3,800 Nvidia Grace Blackwell GPUs in its own European datacenters and runs it there, end to end, under European law. The weights are due as open weights by the end of October.
It doesn't quite match the leading US models, but if you need your data and your provider to stay in Europe, it is the first serious option at this size.
At DevDay on 29 September 2026, OpenAI announced more than 20 new features across ChatGPT, Codex and the API. The common thread: agents that no longer just answer a question but take on ongoing work, operate computers and react to events in other applications – and, if you want, run entirely inside AWS.
