Pakistani artificial intelligence and cybersecurity researchers now have a shot at earning up to $50,000, roughly Rs. 14 million, through a newly expanded bounty program from ChatGPT maker OpenAI.
The company has officially listed Pakistan among the eligible countries on the application page for its OpenAI Bio Bounty Program, a private initiative that pays researchers to probe the biological safety defences of its most advanced AI models.
What the Program Is About
OpenAI has converted what was originally a limited GPT-5.5 Bio Bug Bounty into an ongoing, invitation-only program. The core objective remains unchanged: find a “universal jailbreak” — a single technique or prompt that can consistently bypass the model’s built-in biological safety guardrails across an entire battery of biosafety questions.
Individual workarounds for separate questions will not qualify. OpenAI wants a method that breaks the safeguards systematically, not a collection of one-off tricks.
The findings will be used to patch vulnerabilities before they can be exploited in the real world, particularly in areas where AI could be misused to assist with biological threats.
Reward Doubled
The maximum payout has been raised from $25,000 to $50,000. The full sum is available for a successful universal jailbreak affecting either GPT-5.5 or GPT-5.6. OpenAI has also indicated it may award smaller amounts for partial discoveries at its discretion.
Deadline for GPT-5.5 Testing
Researchers interested in testing GPT-5.5 specifically have limited time. The original bounty scope for that model closes on July 27, 2026. After that date, the program will focus exclusively on GPT-5.6 and future frontier models unless OpenAI announces further changes.
How to Apply
The application process is straightforward. Interested researchers must submit a short form providing their name, country, affiliation, and relevant experience. An active ChatGPT account is also required.
Applications are reviewed on a rolling basis. Those who are accepted will be onboarded to a private testing platform and must sign a non-disclosure agreement before they can begin. Researchers who previously applied to the GPT-5.5 Bio Bounty Program do not need to submit a fresh application.
Not a Public Contest
OpenAI has made clear that this is not an open competition. Submitting an application does not guarantee participation, and only researchers with demonstrated experience in AI red teaming, cybersecurity, biological safety, or related technical fields are likely to be selected.
The program is separate from OpenAI’s other bounty initiatives, which include a Safety Bug Bounty and a Security Bug Bounty covering broader AI safety and conventional cybersecurity vulnerabilities.
For Pakistani researchers who make the cut and manage to crack the biosafety challenge, the program offers both a significant payday and a chance to shape the safety architecture of some of the world’s most powerful AI systems.
