OpenAI rolls out ChatGPT Lockdown Mode: New feature aims to prevent data theft and prompt injection risks

New Delhi:

OpenAI just rolled out Lockdown Mode for ChatGPT for upgraded browsing and securing privacy. As more people use AI in their daily routines, cyber attacks are aiming at these systems keep getting smarter too. So OpenAI came up with this feature to protect users from prompt injection attacks and data theft – stuff that can really put your business or personal info at risk.

They are making Lockdown Mode available across nearly every kind of ChatGPT account – whether you are on the free plan or paying for premium, business, or enterprise access. The point? It helps keep sensitive info from leaking out when you're dealing with risky prompts buried in web pages or files.

What are Prompt Injection attacks?

If someone sneaks any hidden commands into the chatbot and ChatGPT reads them, it can get tricked and could spill some sensitive personal user data – something which should not be shared with a chatbot.

Lockdown Mode does not magically erase these sneaky prompts, but it makes it way tougher for attackers to pull info out of ChatGPT and send it somewhere else.

What changes when Lockdown Mode is enabled?

Many internet-dependent ChatGPT features get activated when a user uses Lockdown Mode.

Some of the key limitations and restrictions include:

• Live web browsing, which stops cached content.
• Deep Research mode will become unavailable.
• Agent Mode is disabled.
• File downloads for data analysis will be blocked.
• External connectors and live integrations will be limited.
• Shopping and financial agent features will be unavailable for the users.

Although these restrictions may reduce some functionality, they could significantly improve security for the users who tend to handle sensitive information.

Features that continue to work

OpenAI has clarified that several popular ChatGPT features remain unaffected by Lockdown Mode.

These include:

• Memory features
• Conversation sharing
• Image generation
• Model improvement preferences
• Manual file uploads
• Codex network access

So you can keep using ChatGPT for lots of things, just within a safer shell.

How to enable ChatGPT Lockdown Mode

Here is how users can activate the feature directly from ChatGPT settings:

• Open ChatGPT.
• Go to Settings.
• Select Security.
• Tap Advanced Security.
• Enable Lockdown Mode.

Keep in mind, though, you cannot use Developer Mode at the same time; turning on one switches off the other.

With updates like these, OpenAI is giving users more control over how ChatGPT touches the outside world. For anyone working with private documents or financial info, that control is not a luxury, rather, it is essential.