Schlechter Einfluss: LLMs können mithilfe versteckter Signale bösartige Merkmale übertragen

I’m guessing that it’s a review of a paper also published in nature on the same date? if so, here’s a link to the original paper that’s not behind a paywall…

[https://www.nature.com/articles/s41586-026-10319-8](https://www.nature.com/articles/s41586-026-10319-8)

The implications are quite profound.

Putting a paywall behind a malicious intent explination of why LLM are doing X is just a paradoxial situation