OpenAI has introduced a Lockdown Mode for ChatGPT that disables web access and certain features to enhance data protection against prompt injection attacks. While this mode improves security, it does not completely eliminate the risk associated with prompt injection, which remains a challenge in the field.