OpenAI has announced the ‘Bio Bug Bounty Program’ for GPT-5.6. Finding a major flaw in its biosafety system will offer a reward of $50,000, double the amount from the previous program. Its goal is to make AI more secure.
ComeArtificial Intelligence (AI) giant OpenAI has made a big announcement for researchers. If anyone finds a major flaw in the biosafety security system of their new GPT-5.6 model, he will get a reward of $ 50,000 (i.e. more than Rs 47 lakh). This announcement has been made under the company’s ‘Bio Bug Bounty Program’. OpenAI previously ran a similar program for the GPT-5.5 model, which has now been extended for GPT-5.6 as well. The special thing is that the maximum amount of reward has been increased from $25,000 to $50,000.
Through this program, OpenAI wants to find ‘universal jailbreak’ methods that can consistently break the biosafety rules of GPT-5.6. This means detecting prompts or techniques that circumvent safety features, not just in a conversation, but in a variety of situations. The company believes that it is important to know such shortcomings in advance to make future AI systems more secure.
The reward will be given not only to those who completely breach the security, but also to the researchers who provide some useful information. The amount of reward will depend on how powerful and effective your discovery is. OpenAI has also clarified that there will be no change in the existing schedule of the GPT-5.5 bio bug bounty program. Researchers can continue testing on that model until July 27, 2026. After this, only GPT-5.6 will be included in this program.
To participate in this Bio Bug Bounty Program, you will have to go to the official website of OpenAI and register on the application page. In the application, you will have to provide information like your name, organization and work experience in this field. The company has said that those applying must have a ChatGPT account. Also, before going to the testing platform, a non-disclosure agreement (NDA) will have to be signed, so that no intelligence information gets out. OpenAI has also said that researchers who have already applied for the GPT-5.5 bug bounty program do not need to apply again.