Oxford researchers awarded funding to investigate hidden attacks on AI systems

The grant will support research into vulnerabilities in vision-language AI agents and how malicious instructions can spread across systems

Raccontata daeng.ox.ac.uk