Okta disclosed a security flaw that allowed unauthorized access to accounts with 52+ character usernames by bypassing password authentication if a stored cache key was detected, affecting users who logged in using the same browser. The vulnerability, introduced in a July 23 update, was only discovered and fixed on October 30, prompting Okta to advise affected customers to review their access logs. Although the issue didn’t impact organizations with multi-factor authentication, the company is urging users to monitor their accounts for any suspicious activity.
Full Article
Anthropic details Constitutional Classifiers, a system that aims to guard AI models against jailbreaks, monitoring both inputs and outputs for harmful content (Cristina Criddle/Financial Times)
Cristina Criddle reports on Constitutional Classifiers, a system developed to protect AI models from jailbreaks by monitoring inputs and outputs for harmful content. Leading tech companies like Microsoft and Meta are also investing in similar safety measures. This new technology aims to enhance the security and reliability of artificial intelligence systems. Full Article
Read more