When prompts were presented in poetic rather than prose form, attack success rates increased from 8% to 43%, on average — a fivefold increase.
The original article found on darkreading Read More
Researchers Use Poetry to Jailbreak AI Models
Related Posts
US Critical Infrastructure Still Struggles With OT Security
How does a company defend itself from cyberattacks by a foreign adversary? A collection of experts gathered at this year’s RSAC Conference to explain how […]
WatchGuard fixes ‘critical’ zero-day allowing firewall takeover
WatchGuard has issued an urgent patch alert for its Firebox firewall appliances after discovering a critical-rated vulnerability that is under exploit by threat actors. Tracked […]
How Gainesville Regional Utilities is locking down vendor risk
Gainesville Regional Utilities (GRU) isn’t just a utilities provider—it’s the communications backbone for the community. In addition to delivering electricity and water, GRU operates fiber-optic […]