From Dark Reading – GitHub’s AI Assistant Opened Devs to Code Theft

Even after a fix was issued, lingering prompt injection risks in GitLab’s AI assistant might allow attackers to indirectly deliver developers malware, dirty links, and more. Read More