Who checks what AI can do?
Science · 20.08.2026 · 393 (6813) · 745-745
DOI: 10.1126/science.ael2161 ↗Kurz gefasst
Editorial in der Zeitschrift Science, erschienen am 20.08.2026. Verfasst von Thorsten Holz. Beteiligte Einrichtung: Thorsten Holz is a scientific director at the Max Planck Institute for Security and Privacy, Bochum, Germany.
Abstract (Originalsprache)
The most important findings about frontier artificial intelligence (AI) are also the hardest to verify. Much of the information needed to understand its capabilities and risks—including results from evaluations of prerelease models and containment experiments—remains largely inaccessible outside the labs that produce it. In recent weeks, OpenAI, Anthropic, and Meta disclosed that research models had reached beyond their intended testing environments and compromised other organizations’ systems. Those labs deserve credit for reporting this. But outside those labs, there was no way to discover, reproduce, or verify what had happened.
Autor:innen und Institutionen
- Thorsten Holz is a scientific director at the Max Planck Institute for Security and Privacy, Bochum, Germany.
Zitieren
APA
Holz, T. (2026). Who checks what AI can do?. Science, 393(6813), 745-745. https://doi.org/10.1126/science.ael2161
Vancouver
Holz T. Who checks what AI can do?. Science. 2026;393(6813):745-745. doi:10.1126/science.ael2161
BibTeX
@article{Holz2026,
author = {Holz, Thorsten},
title = {Who checks what AI can do?},
journal = {Science},
year = {2026},
volume = {393},
number = {6813},
pages = {745--745},
doi = {10.1126/science.ael2161}
}
EUVIMED