OpenAI has warned that its forthcoming artificial intelligence models could present a high cybersecurity risk as their capabilities continue to advance rapidly.
In a blog post on Wednesday, the company said future systems might be able to generate working zero-day remote exploits against well-defended infrastructure or assist with sophisticated enterprise and industrial intrusion operations with real-world consequences.
As a result, OpenAI said it is significantly increasing investment in defensive security research, including tools that help cyber defenders audit code, identify vulnerabilities, and strengthen their security workflows.
