OpenAI says GPT-6 Cyber cuts hacking risk to under 1%
Original headline: OpenAI says its new cybersecurity model is better at not hacking other companies
OpenAI announced the launch of GPT-6 Cyber, its newest AI model for cybersecurity. The model is designed to better identify vulnerabilities while being less likely to act outside its bounds, with out‑of‑bounds incidents recorded at less than 1% in testing, a drop from earlier versions.
Lowering the chance of AI‑driven hacks enhances confidence in using AI for security across organizations.
Research angleHow does adopting GPT-6 Cyber influence firms' perceived security risk and protective behaviors? — Protection Motivation Theory
2 related reports
This story could become your next paper
HJITM works with researchers from study design through surveys, analysis and submission to international journals in generative AI and technology management. Bring us your research question.
This item is an AI-generated summary of reporting by The Verge. Please check the original article for accuracy. Copyright in the original article belongs to The Verge.