OpenAI Launches Lockdown Mode to Enhance ChatGPT Security Against Cyberattacks
OpenAI has introduced a new feature called Lockdown Mode for its ChatGPT users, aimed at enhancing security against cyberattacks, specifically 'prompt injection' attacks. These attacks involve embedding malicious instructions within web pages and other content sources that could potentially compromise the chatbot's functionality. Lockdown Mode restricts several system capabilities, including disabling live web browsing and limiting the use of cached data. It also suspends functions like searching and displaying images from the internet, as well as Deep Research and Agent Mode. Despite these restrictions, users can still generate images. OpenAI acknowledges that while Lockdown Mode increases security, it does not guarantee complete protection, as cached content or uploaded files might still harbor malicious commands. The feature is primarily targeted at organizations and individuals dealing with highly sensitive data, and it is being gradually rolled out for ChatGPT Business and select personal accounts. 
