Factoring RSA Keys with Many Zeros

A collection of security items: new research finds many real-world RSA public keys with unusually sparse (many-zero) moduli that are factorable with specialized techniques; an open dataset (badkeys) found such weak keys in the wild. A large leak exposed nearly one million passports, showing how high-value credentials can be compromised via low-value systems. Multiple AI risks: prompt-injection research demonstrating model role-confusion, malware authors embedding policy-triggering “forbidden” text to thwart AI analysis, and a rapid jailbreak of Anthropic’s Fable 5 safety guards. Privacy and ML