AI Leaders Urge Mandatory DNA/RNA Screening to Reduce AI-Enabled Biosecurity Risks

Image checked for display before being inserted into this Markdown file. Image extracted from WIRED, JPG format, not SVG. WIRED credits it as “Photo-Illustration: Skye Battles; Getty Images.”¹

Quick summary

A group of AI leaders, biology experts, national security specialists and gene synthesis companies has signed an open letter urging the U.S. Congress to require mandatory synthetic DNA/RNA order screening and related recordkeeping. Signatories include Sam Altman of OpenAI, Dario Amodei of Anthropic, Demis Hassabis of Google DeepMind and Mustafa Suleyman of Microsoft AI.²³⁴

The core message is that AI is advancing quickly and may reduce the knowledge barriers that historically made biological misuse difficult. At the same time, synthetic DNA/RNA can be ordered online for legitimate research. The letter argues that providers should screen orders, verify customers and keep records to reduce misuse risk.³²

This is notable because leading AI rivals rarely agree publicly on regulation, but they are aligned here on a specific biosecurity safeguard.

What happened?

The open letter is titled “In Support of Mandatory Nucleic Acid Synthesis Screening and Recordkeeping.” It calls on legislators to require companies that sell synthetic DNA/RNA and related synthesis equipment to check whether orders contain concerning sequences and to verify that customers are legitimate before fulfilling them.³

WIRED reported that leaders from OpenAI, Anthropic, Google DeepMind and Microsoft AI are among the signatories. The letter was organized by the Institute for Progress and the Foundation for American Innovation.²

The Verge also reported that the letter represents a rare moment of agreement among major AI competitors around stronger biosecurity safeguards.⁴

Who signed the letter?

The signatories include several groups:

The letter’s website lists Demis Hassabis, Sam Altman, Dario Amodei, Alexandr Wang, Mustafa Suleyman and many others among the signatories.³

Why this is becoming a major AI story

AI can help biological research, drug discovery and protein design. That is the upside. But it can also make complex technical knowledge more accessible. Experts worry this could reduce the barriers for malicious actors in biology.²³

The letter does not call for banning biology research. It explicitly recognizes that synthetic DNA has helped vaccine development, basic research and diagnostics. The issue is the need for a control point in the biotechnology supply chain to reduce misuse.³

What is synthetic DNA/RNA screening?

In this context, synthetic DNA/RNA screening means that providers check:

• whether the customer is legitimate;
• whether an order contains concerning sequences;
• whether an order is related to dangerous pathogens or toxins;
• whether records are sufficient for investigation if a biosecurity incident occurs.

Many major companies already do some screening voluntarily. WIRED notes that the International Gene Synthesis Consortium was formed in 2009 to implement voluntary safeguards in the gene synthesis industry.²

The new proposal is to move from voluntary screening toward a mandatory legal requirement.

Why AI makes the issue more urgent

The open letter says the pace of AI progress is the new factor. It argues that AI systems could erode the knowledge barriers that previously made biological misuse harder.³

WIRED also quotes experts warning that AI can help users find providers or reframe requests in ways that may make screening harder. The article emphasizes that screening alone is not enough; AI companies also need their own safeguards and monitoring.²

The issue is not simply “AI is dangerous.” It is that AI and synthetic biology are advancing together, and biosecurity governance needs to keep up.

What else are AI companies doing?

This news fits a broader pattern: AI companies are increasingly using controlled access programs for models with sensitive capabilities.

The Wall Street Journal recently reported that OpenAI created a limited-access program for a science-focused model called GPT-Rosalind, aimed at helping government agencies and trusted public health organizations address biological risks such as pandemics and biological weapons.⁵

Reuters also reported that the U.S. government expanded its risk-assessment program for unreleased AI models from Google DeepMind, xAI and Microsoft. The reviews focus on risks such as cyberattacks, chemical/biological weapons and training data integrity.⁶

The common trend is that AI labs are accepting more testing, monitoring and controlled distribution for models with high-risk capabilities.

Impact on the AI industry

This news may affect the AI industry in several ways:

1. AI safety becomes more concrete

Instead of debating only long-term AGI risk, this proposal focuses on a specific supply-chain control: DNA/RNA order screening.

2. Major AI rivals have found common ground

OpenAI, Anthropic, Google DeepMind and Microsoft AI are usually competitors, but they are aligned on this issue. That strengthens the policy signal.

3. Synthetic biology providers may face new rules

If legislation passes, U.S. providers of synthetic DNA/RNA may need stronger screening and recordkeeping. That could raise compliance costs, but may also increase trust in the industry.

4. AI-for-biology companies will need stronger compliance

Biotech, pharma, lab automation and AI-for-science companies may need clearer processes for ordering, recordkeeping and demonstrating legitimate use.

Impact on developers and AI businesses

For developers and companies building AI products, this story signals that not every AI capability should be scaled without access controls.

Practical lessons:

• high-risk AI domains need access control;
• logging and audit are becoming mandatory features;
• user verification matters for specialized tools;
• guardrails need operational monitoring;
• models powerful in biology, cybersecurity or chemistry may need tiered access;
• compliance may become a competitive advantage.

Open questions

Several issues remain unresolved:

• which companies will be covered by mandatory screening;
• how compliance costs affect small biotech startups;
• how long order records will be stored and protected;
• whether customer verification slows legitimate research;
• whether non-U.S. providers adopt similar rules;
• how much screening AI labs should do for bio-related prompts;
• how to balance biomedical innovation and biosecurity.

Conclusion

The letter from major AI leaders calling for mandatory synthetic DNA/RNA screening is an important moment in AI safety policy. The proposal is concrete: screen orders, verify customers and keep records at a key point in the biotechnology supply chain.

If implemented carefully, it could reduce the risk of AI-assisted biological misuse while preserving the benefits of synthetic biology for medicine, vaccines and research. It also shows that the next phase of AI governance will increasingly focus on the intersection of AI, biology, national security and high-tech supply chains.

FAQ

What are AI leaders asking for?

They are asking U.S. lawmakers to require synthetic DNA/RNA order screening, customer verification and recordkeeping to reduce misuse risk.³²

Who signed the letter?

Signatories include leaders from OpenAI, Anthropic, Google DeepMind, Microsoft AI, Meta, Stripe, synthetic biology companies and many biosecurity experts.³

Why is AI relevant to biosecurity?

AI can make complex technical knowledge more accessible, potentially lowering barriers for biological misuse.²⁴

Does the proposal ban biology research?

No. It focuses on screening orders and verifying customers, not banning legitimate research.³

Why is this important for AI policy?

It is one of the clearest examples of AI safety moving toward specific, enforceable policy in a high-risk domain.

References

Footnotes

1. WIRED. “OpenAI and Anthropic Sign Letter to Prevent AI-Developed Biological Weapons.” Image: Photo-Illustration: Skye Battles; Getty Images. https://media.wired.com/photos/6a1f6447234d4b89dad80277/1%3A1/w_2560%2Cc_limit/science_anthropic_final.jpg ↩

2. WIRED. “OpenAI and Anthropic Sign Letter to Prevent AI-Developed Biological Weapons.” https://www.wired.com/story/openai-anthropic-letter-ai-biological-weapons/ ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

3. “An Open Letter in Support of Mandatory Nucleic Acid Synthesis Screening and Recordkeeping.” https://screendna.org/ ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰

4. The Verge. “AI leaders call for tougher protections against AI-aided bioweapons.” https://www.theverge.com/ai-artificial-intelligence/942956/ai-biological-weapons-open-letter-congress ↩ ↩² ↩³

5. The Wall Street Journal. “OpenAI Opens Up Use of AI Model to Address Biological Risks.” https://www.wsj.com/livecoverage/stock-market-today-dow-sp-500-nasdaq-05-29-2026/card/openai-opens-up-use-of-ai-model-to-address-biological-risks-bRcAyBPWSl13DmHkvjH2 ↩

6. Reuters. “What we know about US stress tests of Google, xAI and Microsoft AI models.” https://www.reuters.com/legal/litigation/what-we-know-about-us-stress-tests-google-xai-microsoft-ai-models-2026-05-05/ ↩