Archive
Security research highlights how large language models could exploit parsing flaws in inference software like vLLM to execute code and compromise high-value data center servers.
San Francisco-based AI assistant startup Instinct is generating buzz for its task automation, but early testing has sparked severe privacy and security concerns over how the platform handles user data.
A Connecticut state court judge has issued sanctions against a self-represented plaintiff who hid invisible AI prompt instructions inside legal filings, marking what appears to be the first documented instance of a US litigant attempting to manipulate judicial software.
