The new exercise follows an earlier security drive in which OpenAI assigned 25 percent of its product engineers to finding and fixing vulnerabilities.
OpenAI has committed $1 billion (roughly Rs. 9,500 crore) to support cyber defenders
Photo Credit: Unsplash
OpenAI is working with security firm Cathedral to find weaknesses in its systems. Researchers from both teams will use OpenAI's AI models to test its defences and check whether potential fixes work. The company is also raising rewards for critical vulnerabilities reported through its Security Bug Bounty programme by more than four times. The latest exercise follows an earlier round of security testing in which OpenAI assigned 25 percent of its product engineers to finding and fixing vulnerabilities.
Cathedral's researchers will join OpenAI's teams to look for security flaws in the company's systems, the company said in a blog post. They will use OpenAI's AI models to test its defences and see if potential fixes work. Gavin Kliger, CEO and co-founder of Cathedral, said the partnership would bring the firm's security researchers together with OpenAI's teams to find weaknesses that might otherwise go unnoticed.
OpenAI is also increasing the rewards offered to researchers who report serious security flaws through its bug bounty programme. The company has not disclosed the new payout amounts, but says rewards for critical findings will be more than four times higher than before.
The new exercise follows an earlier security drive in which OpenAI assigned 25 percent of its product engineers to finding and fixing vulnerabilities. They used Codex to investigate reported problems, recreate them in isolated environments and prepare fixes for review.
OpenAI is also giving Daybreak customers access to its AI models. The company has committed $1 billion (roughly Rs. 9,500 crore) in subsidised access over six months to help organisations protecting essential services and digital infrastructure use AI in their cybersecurity work.
The company plans to share what it learnt from the previous security exercise, including how it was organised and the tools and processes involved. It will also publish a guide for other companies and government agencies looking to use AI to improve their cybersecurity.
Get your daily dose of tech news, reviews, and insights, in under 80 characters on Gadgets 360 Turbo. Connect with fellow tech lovers on our Forum. Follow us on X, Facebook, WhatsApp, Threads and Google News for instant updates. Catch all the action on our YouTube channel.