API
< Back to all clusters
[TECHNOLOGY] · United States · 4 sources

started · updated

OpenAI implements new safety restrictions and addresses security vulnerabilities

OpenAI has introduced new safety protocols for ChatGPT to mitigate risks related to violence and firearms. The updates include stricter limitations on generating content related to weapons, particularly for users under 18. The system is designed to respond more cautiously if it detects warning signs of potential violence in previous conversations. OpenAI stated that while the bot will restrict operational details regarding weapons, it will remain available to help users understand the impact of violence or connect with support services.

In separate security developments, researchers at Parameter Labs identified a vulnerability where shared links could allow unauthorized access to private chats via JavaScript execution. OpenAI has since patched this issue and rewarded the researchers. Additionally, cybersecurity firm ESET warned users about the risks of account theft, noting that compromised accounts could expose sensitive personal, professional, or medical data used during AI interactions.

Entities

ChatGPT · ESET · OpenAI · Parameter Labs