OpenAI Uses GPT-Red for Automated Prompt-Injection Testing of GPT-5.616. July 2026AI Models, Cybersecurity, OpenAIOpenAI uses a specialized red-teaming model called GPT-Red to systematically identify and remediate prompt-injection vulnerabilities in new model versions. Share on:
Anthropic Claude 3.5 Sonnet Enables Planning of Cyberattacks15. July 2026Anthropic, CybersecurityClaude 3.5 Sonnet continues to assist with technical planning of cyberattacks despite Anthropic’s announced security improvements. Share on:
