OpenAI introduces a bug bounty program to improve the platform security

The biggest challenge to any company is the software vulnerabilities and bugs. These problems couldn’t be neglected as hackers could gain access to sensitive data and information. In order to cater to this problem, OpenAI has introduced an interesting program. This program is named the Bug Bounty program. Through this program, the users who identify security vulnerabilities will be rewarded. The Bugcrowd platform will be used for the bounty program, which will give out cash incentives of up to $20,000 for extraordinary discoveries and $200 for less severe ones.

OpenAI declares that the main target area of this program is privacy and cybersecurity in the company’s web domains and APIs. In addition to this, the firm is also interested to rule out the weaknesses in data processing by ChatGPT. It will even provide third-party access to it via its API.

Researchers can use the Bugcrowd platform to report bugs in the API and ChatGPT. If a model flaw does not have a security impact, OpenAI has asked researchers to report it using a separate form. Model safety issues are not discrete bugs that OpenAI can easily fix; rather, they require extensive research and a more comprehensive approach.

Constraints of the Bug Bounty Program

The bug bounty program by OpenAI has some limitations. Jailbreaks for ChatGPT or text cues meant to deceive the AI program into breaking its own rules won’t be accepted by OpenAI. Additionally, the business will not accept reports of ChatGPT misbehaving or fabricating information in accordance with the program.

On the other hand, Apple and Google pay up to $2 million for the most severe vulnerabilities. In comparison, the reward amount by OpenAI is much less. Still, it provides the users with a chance to help improve the platform. In this way, they themselves and others are protected from security threats.

“We invite you to report any bugs, security holes, or vulnerabilities you find in our systems. In this way, you will play a critical part in making our technology safer for everyone by sharing your results, claims OpenAI.