Source: The Hacker News
Author: info@thehackernews.com (The Hacker News)
URL: https://thehackernews.com/2026/06/new-chatgpt-lockdown-mode-limits-tools.html
https://thehackernews.com/2026/06/new-chatgpt-lockdown-mode-limits-tools.html
ONE SENTENCE SUMMARY:
OpenAI’s ChatGPT Lockdown Mode reduces prompt-injection data exfiltration risk by restricting networked tools, while adding session management controls.
MAIN POINTS:
- Introduces optional Lockdown Mode for eligible personal accounts to mitigate prompt-injection exfiltration.
- Targets users handling sensitive data needing stronger protection guarantees.
- Available across Free, Go, Plus, Pro, and self-serve Business plans.
- Limits tools connecting to web or external services to reduce outbound data leakage.
- Builds on sandboxing and controls against URL-based exfiltration techniques.
- Focuses on removing exfiltration pathways, not preventing prompt injections outright.
- Leaves memory, file uploads, and conversation sharing behavior unchanged.
- Disables or restricts browsing, images, deep research, agent mode, canvas networking, and downloads.
- Mutually exclusive with Developer Mode; enabling one automatically disables the other.
- Adds session review/logout feature with device, app, location, timing, and trust indicators.
TAKEAWAYS:
- Activate Lockdown Mode when sensitive data exposure would be high impact.
- Expect reduced functionality as a tradeoff for fewer outbound exfiltration routes.
- Recognize residual risk from apps, capability combinations, or novel techniques.
- Understand prompt injections can still manipulate outputs even without data theft.
- Use new session-management tooling to detect and respond to account compromise quickly.