OpenAI Introduces Lockdown Mode to Enhance ChatGPT Security Against Cyberattacks
OpenAI has launched a new feature called Lockdown Mode for its ChatGPT users, aimed at bolstering security against cyberattacks, specifically 'prompt injection' attacks. This mode restricts several system capabilities, including live web browsing, and limits the use of cached data to prevent malicious instructions from affecting the chatbot. Despite these measures, OpenAI acknowledges that complete security cannot be guaranteed, as cached content or uploaded files may still pose risks. The feature is primarily targeted at organizations and individuals handling sensitive data, and is being gradually rolled out for ChatGPT Business and select personal accounts. 
