Pakistani researchers can apply for OpenAI’s $50,000 bio safety bounty

Pakistani researchers specialising in artificial intelligence security and biosafety can apply for a reward of up to $50,000 under a private bounty programme launched by OpenAI.

Pakistan is included among the selectable countries on the programme’s application form, allowing eligible researchers based in the country to seek participation. However, submitting an application does not guarantee acceptance, as OpenAI will select participants for the private initiative.

The OpenAI Bio Bounty Program is aimed at identifying weaknesses in safeguards designed to prevent advanced AI models from being misused for biological risks.

Selected researchers will be required to find a universal jailbreak capable of consistently bypassing the model’s biological safety protections across a predefined set of questions. Separate workarounds for individual prompts will not qualify as a universal jailbreak.

OpenAI has doubled the programme’s maximum reward from $25,000 to $50,000. The full amount may be awarded for a successful universal jailbreak affecting GPT-5.5 or GPT-5.6, while smaller rewards may be offered for partial findings.

Testing involving GPT-5.5 will remain open until July 27, 2026. After that date, the programme will focus on GPT-5.6 and future frontier models unless the company announces further changes.

Applicants must provide their name, country, affiliation and relevant experience, and must already have a ChatGPT account. Those selected will be required to sign a non-disclosure agreement before receiving access to the private testing platform.

The opportunity is intended mainly for researchers with experience in AI red teaming, cybersecurity, biological safety and related technical fields rather than the general public.