Artificial intelligence company OpenAI has taken drastic steps to find vulnerabilities in its technology systems.
Now they will offer up to 20,000 dollars, equivalent to 210,000 kroner, to people who report “bugs”, writes Reuters. The company is behind the famous text bot ChatGPT.
“Security is essential to OpenAI’s mission. We appreciate the contributions of ethical hackers who help us maintain high standards of privacy and security for our users and technology,” wrote the company when they launched the project, which they call Bugcrowd.
Blocked in Italy
Just under two weeks ago, the Italian authorities blocked ChatGPT due to what they believe is a breach of EU privacy rules. They also believe the robot does not verify the users’ age.
Other countries have since started their own investigations to check whether ChatGPT follows the rules for privacy.
OpenAI writes that they have also invited researchers to review certain functions of the text robot, as well as look at the framework for how the systems communicate and data sharing with third-party applications.
14 weaknesses found
The amount for any errors will vary according to severity. The minimum payment is 200 dollars per errors that are found.
14 vulnerabilities have been found so far according to OpenAI, which has paid an average of $1,287.50 for each of them.
However, the artificial intelligence platform makes it clear that they do not reward people who get the bot to provide wrong information or say malicious things. You also get no profit for jailbreaking, i.e. removing the restrictions, for example, that Apple has set for its operating system.
