Deepfakes, TikTok & teenagers: what the new VAMPIR research project aims to change

← All articles
Artificial intelligence · EN

Deepfakes, TikTok & teenagers: what the new VAMPIR research project aims to change

Video‑based content dominates adolescents’ media use. At the same time, deepfakes and video‑based disinformation are becoming more realistic, faster to spread, and harder to detect (especially on platforms such as TikTok, Instagram and YouTube). This poses particular risks for young users: manipulated videos can influence opinions, exploit emotions, and obscure underlying interests, funding sources or political agendas.

Against this backdrop, a new interdisciplinary German research project has started in April 2026: VAMPIR – “Video Use in Adolescence: Manipulation, Propaganda, Information and Resilience”. The project is funded by the Federal Ministry of Research, Technology and Space (BMFTR) with €2.07 million and runs until March 2029.

What makes VAMPIR distinctive?

The project brings together computer science, psychology, media studies and law to address videobased manipulation from multiple angles:

1) Technical: development and assessment of AI‑based methods to detect deepfakes and video disinformation.
2) Psychological: analysis of how video‑based manipulation differs from text‑based disinformation and how adolescents process and respond to such content.
3) Media & communication studies: examination of dissemination dynamics on platforms and the role of influencers.
4) Legal: evaluation of whether current EU and national frameworks are fit to protect minors (and where gaps remain).

VAMPIR examines, among other things:

1) data access for researchers under the Digital Services Act (DSA);
2) youth‑specific risks under the GDPR, including profiling, targeting and AI training with minors’ data;
3) platform obligations under the DSA, AI Act, AVMS Directive, and national youth‑protection regimes;
4) duties to assess and mitigate systemic risks (Art. 34 DSA);
5) labelling obligations for synthetic content and deepfakes (Art. 50 AI Act);
6) the interaction between state regulation, platform governance and controlled self‑regulation (hybrid governance).

The regulatory debate increasingly targets platform risks, AI‑generated content and the protection of minors. VAMPIR’s added value lies in connecting empirical evidence with concrete legal and policy recommendations.

I added the reference to highlight that influencer‑driven disinformation poses a distinct risk for adolescents, particularly where underlying economic incentives are not transparently disclosed.

This post has a document attached on LinkedIn. View on LinkedIn →

Discuss on LinkedIn