Niklas Titgemeyer. DE
Niklas Titgemeyer in a blue suit

Niklas TitgemeyerContent creator & expert on
risks from AI chatbots

Request a keynote
+0
Followers across platforms
+0
Years of experience in AI.Years of
AI experience
+0
Years of experience in journalism.Years in
journa­lism
(television and newspaper)
0%
of all claims backed by sourcesBacked by
sources

These partners trust my expertise

Keynote speaker with a focus
on the risks of AI

Niklas Titgemeyer is a keynote speaker and sociologist (M. A.). His specialisms are the risks and limitations of AI. His goal is to educate audiences on how to use chatbots like ChatGPT in a critical and responsible way.

His talks look behind the curtain of the most widely used language models. They cover the limitations and risks in the practical use of AI, as well as what these technologies mean for democracy and society, especially with regard to misinformation and cognitive autonomy. With more than 125,000 followers on social media, he is one of the most influential voices on the limitations of language models in the German-speaking world.

His talks are delivered in English or German, on site or online, and are tailored to your industry.

Book an AI keynote as the
highlight of your event
Niklas Titgemeyer giving a talk

Choose from four AI keynotes

Every keynote is tailored to your industry. The four topics below serve as the foundation, because they are where Niklas Titgemeyer's expertise sits.

Symbol for the limits and risks of language models.

The risks of working and gathering information with chatbots

Language models sound remarkably intelligent. Not only when their answers are correct, but also when they are wrong. However, according to research, they are wrong far more often than most people assume. In this keynote, Niklas Titgemeyer shows the real dangers of working with language models like ChatGPT.

Book this keynote
A portrait with a digital glitch effect across the face, symbolising AI and disinformation.

A crisis of trust: how deepfakes and AI propaganda put us to the test

Deepfakes, synthetic images and AI-generated text at scale are changing how disinformation is produced and how it spreads. For public debate, that is not good news. It leads to a disruption of trust, which is essential for democracies. This keynote shows how AI-driven disinformation works and what problems it creates.

Book this keynote
A person sitting alone at night in front of a laptop and a monitor, typing into an AI application.

Dangerous trust: when ChatGPT becomes a therapist

More and more people talk through their personal crises with AI chatbots. For technical reasons, however, these models tend to agree with their users instead of pushing back. What that does to people in acute distress is documented in cases of AI-induced psychosis, but also in studies that explain the mechanism behind it.

Book this keynote
Servers in a data centre, lit in blue and red, symbolising increasingly powerful AI systems.

Why advanced AI chatbots are an existential threat

The very people building these systems are now the ones warning most loudly about them. The more capable the models become, the harder they are to control. In this keynote, Niklas explains the arguments behind AI doomsday scenarios and why we should at least take them seriously.

Book this keynote

Book Niklas Titgemeyer as a keynote speaker for your event.

You will receive a non-binding offer within 24 hours of your enquiry.

Request a keynote

Expert appearances by
Niklas Titgemeyer

Media outlets regularly ask Niklas Titgemeyer to appear as an expert on AI and its risks. Here is a selection of his appearances (in German).

9:41
TV feature Brisant, 29 September 29 Sep 2026 · BRISANT +++ OpenAI halts new AI version Brisant is produced by Mitteldeutscher Rundfunk (MDR) and has aired daily at 5 pm on Germany’s Das Erste since 1994.
9:41
Why you should never blindly trust artificial intelligence WUNDERPANIK
9:41
generation:ki #03 | Human. Machine. Together? How do we want to live with AI?
Spotlight Medienbildung2.8K subscribers
16:30 We are not ready for the deepfake revolution.Niklas Titgemeyer
1/3

Assessing the risks of new AI models on Brisant (ARD)

On 29 September 2026, the German TV magazine Brisant (ARD) reported that OpenAI had halted a new AI version. Niklas Titgemeyer appeared in the feature as an expert. In broad terms, he says there are three reasons why news like this is becoming more and more common.

  1. Deception: For some time now, AI agents have tended to deceive their users under certain conditions. For example when they cannot find a solution or have made a serious mistake. A year ago there were already cases in which agents deleted entire databases and then concealed it.
  2. Hacking skills: Current models find vulnerabilities even in the most modern software of large companies and are able to exploit them. On its own, that is not necessarily a problem.
  3. Training for persistence: It becomes a problem with the third factor. Models are currently being trained to complete ever longer tasks and not to give up. If they cannot find a solution, they have to be trained not to take the shortcut in that case (in other words, not to use their hacking skills).

So it is understandable that, given their abilities, it is becoming ever harder to teach newer model generations to work on a task for a long time and try all kinds of things without falling back on their hacking skills. For the alignment of these models (that is, aligning them with human values and goals), this is a major challenge.

In a podcast interview on the limitations of AI language models

In the podcast “WUNDERPANIK”, Niklas Titgemeyer talks with Katharina Mager‑Micijevic and Melanie Hölting‑Eckert about the two biggest risks of working with AI chatbots.

How often they tend to generate false information (hallucinations), in which fields and use cases this happens most, and why they are so convincing when they simply tell us what we want to hear. Because AI models mainly tell us what makes us feel good, not necessarily what is true. They are trained to generate answers that people prefer.

On difficult topics such as medicine or recent research, AI even hallucinates in 30 to 60 % of cases. Smaller models are worse still and reach up to 85 %. This also applies to the newest and most capable models such as Fable 5, which reach 50 % hallucinations on medical or legal questions without web search and around 40 % with web search.

The interview also asks why outsourcing our thinking to AI is a problem, and under which conditions it is in fact reasonable to trust a chatbot blindly, even if that sounds counterintuitive at first.

Here’s the episode (in German)

Using AI responsibly: an expert interview with “Digitale Helden”

The team at Digitale Helden supports schools and families with online media education, so that young people can navigate the digital world safely, independently and with confidence.

Their project “generation:ki” has produced a feature on the question “How do we want to live with AI?”. Niklas Titgemeyer is one of seven experts in it and answers questions such as:

  • How does AI actually work?
  • How long have AI technologies been around?
  • What risks does AI bring with it?
  • How do you use AI responsibly?
  • Which limitations should you keep in mind?
  • What should young people who are just starting out with AI pay attention to?
Watch the interview (in German)

What your audience
takes
away from
an AI keynote

  1. 1

    Spotting risks in daily work: Your audience learns how language models mislead us with their answers, why they invent information and why they tell us what we want to hear.

  2. 2

    Safer and more efficient use: A keynote gives people the ability to judge AI chatbots properly and to build a critical habit of working with them.

  3. 3

    Learning that lasts: Detailed animations and graphics turn complex topics into images, so that they stay with your audience long after the event.

  4. 4

    Critical reflection: After the talk, your audience will rethink how they have been using AI, because they have genuinely learned something.

An audience at a talk, seen from behind, facing the screen.

Why book Niklas Titgemeyer
as an AI speaker?

  • A rare combination: trained sociologist (M. A.), journalist and expert on the risks and weaknesses of language models
  • Understanding how technology and society interact
  • Understanding the risks and limitations of AI
  • Presentations tailored to your industry
  • Graphics and animations instead of text-heavy slides
  • Delivered in English or German, on site or online
Request a keynote

Sociologist

AI is not only changing the way we work, it is changing the way we live together. As a sociologist, Niklas Titgemeyer is particularly interested in what generative AI does to society, democracy and individuals. Its sudden success has left visible marks in a very short space of time.

Journalist

More than four years in journalism (one year as a working student in television, then three years as a freelance reporter for a daily newspaper) still shape his work today. That means researching thoroughly and transparently, and then presenting a topic in a way that is both accurate and genuinely understandable.

LLM expert

For Niklas Titgemeyer, expertise in language models means more than practical experience. It means knowing how these models actually work. Over several years of formal training and reading a large volume of AI research, he has built a technical understanding of language models that lets him explain the engineering side as well.

Portrait of Niklas Titgemeyer More about Niklas Titgemeyer,
his background and his expertise

Frequently asked questions

Why book a keynote speaker for your event?

With topics like AI and digitalisation, a good keynote speaker removes the fear of touching the subject, opens up new perspectives and leaves an audience that wants to act.

How long is a talk?

The length adapts to your event format. It ranges from 20-minute impulse talks to full 90-minute keynotes.

Are examples from my industry included?

Yes. Every talk is tailored to the specifics and requirements of your organisation and your industry.

Can the keynotes be delivered remotely?

Yes. Talks can be delivered on site or as an online format.

How far in advance should I book?

About two months of lead time is recommended, so the talk can be prepared properly for your audience.

In which languages does Niklas Titgemeyer speak?

Talks are delivered in English or German.

Where is Niklas Titgemeyer based, and does he travel?

He is based in Germany and travels for events across Europe.

Book an AI keynote as the
highlight of your event

Fill in the form and book
Niklas Titgemeyer as your AI
keynote speaker

By submitting this form you consent to your details being processed in order to handle your enquiry. More on this in the privacy policy.

Book a free discovery callFree call (15 min)