When data poisoning in AI models becomes a security challenge
Manipulated training data can deliberately alter the behavior of AI systems – often without anyone noticing. According to cybersecurity expert Pierre Delcher, this raises a strategic question for Europe: How can we protect the integrity of digital knowledge systems before disinformation becomes part of the infrastructure itself?
PD
26. May 2026
26. May 2026
Last updated: 26. May 2026
Share article: