OpenAI has rolled out new security features for ChatGPT, including „Lockdown Mode“ and „Elevated Risk“ labels. These additions aim to bolster organizational defenses against prompt injection attacks and potential AI-driven data exfiltration.
Source: OpenAI