OpenAI Rosalind Biodefense and GPT-Rosalind Expansion

OpenAI has launched the Rosalind Biodefense initiative and expanded trusted access to GPT-Rosalind, a frontier reasoning model built for life sciences research, to vetted developers and government partners. These steps aim to provide defenders with advanced AI tools to prevent, detect, and respond to biological threats more effectively than those who might misuse the technology.

Rosalind Biodefense Initiative for Trusted Developers

OpenAI is providing sponsored access to GPT-Rosalind and launch support to trusted developers building high-impact defensive applications in the life sciences. The program focuses on operationalizing frontier AI capabilities into practical biodefense tools to strengthen societal resilience before biological threats emerge.

Targeted areas for development include:

• Epidemiological modeling
• Early detection and screening
• Preparedness and response capabilities
• Non-pharmaceutical interventions (NPIs)
• Other public-health-relevant capabilities

OpenAI is specifically seeking projects that improve the speed, quality, or scale of defensive research workflows, such as literature synthesis, protocol design, data harmonization, simulation, and decision support. The program is open to academic, nonprofit, government-affiliated, and mission-driven companies.

Early Adopters and Use Cases

One of the first organizations supported by the program is Fourth Eon Biosecurity, which is developing AI-native biosecurity screening systems. These systems analyze sequences to generate threat assessments and provide function-based screening for DNA synthesis to prevent unsafe or malicious orders.

"We’re excited to test OpenAI†s GPT-Rosalind in Fourth Eon’s work developing AI-native biosecurity screening systems that analyze sequences and generate detailed threat assessments. Robust screening can improve the ability to detect and mitigate potentially dangerous DNA orders before they create downstream risk, strengthening prevention."

† — Gary Abel, Co-Founder & Chief Scientist, Fourth Eon Biosecurity

Expanded Trusted Access for Government and Allied Partners

OpenAI is extending trusted access to GPT-Rosalind for select U.S. government and allied partners with approved public health and biodefense missions. This model allows qualified teams to apply frontier AI to high-impact workflows while maintaining safety, security, and accountability controls.

Key applications for government partners include:

• Early warning systems
• Outbreak response planning
• Diagnostics and preparedness
• Medical countermeasure development

Strategic Partnerships

Several institutions are already utilizing GPT-Rosalind for critical biodefense missions:

• Lawrence Livermore National Laboratory (LLNL): Integrating AI with supercomputing and simulation to design and evaluate medical countermeasures for emerging biological threats.
• Johns Hopkins Applied Physics Laboratory: Integrating the model into a protein-engineering platform to accelerate the screening of mutant enzymes for therapeutics and biothreat characterization.
• Coalition for Epidemic Preparedness Innovations (CEPI): Utilizing the model to support the "100 Days Mission" to accelerate vaccine development against pandemic threats, including the current Ebola outbreak.

Safety Framework and Layered Resilience

These initiatives build on a broader safety strategy to ensure AI capabilities in biology are deployed responsibly. OpenAI's approach involves "layered resilience," which includes:

• Preparedness evaluations and bio-specific capability assessments
• Safer model behavior for dual-use biological requests
• Monitoring, enforcement, and expert red teaming
• Security controls for high-risk capabilities

In July 2025, OpenAI identified the ChatGPT agent as the first model treated as "High Capability" in biology under its Preparedness Framework, leading to the activation of robust safeguards. The current expansion of GPT-Rosalind access is the result of ongoing collaboration with external testing groups and public-sector partners, including the U.S. Center for AI Standards and Innovation (CAISI), the UK AI Security Institute (UK AISI), Los Alamos National Laboratory, and the Frontier Model Forum.