[This project was initially proposed through Lightcone Commons, however we are still waiting for a funding decision and seeking additional support would be welcomed!]

Project summary

We are seeking $206,000 - $420,500 USD to support the next year of development of Commec (the Common Mechanism), the free, open-source, globally-available DNA synthesis screening tool hosted by IBBIS, the International Biosecurity and Biosafety Initiative for Science.

Why fund DNA synthesis screening tools?

• Safeguard a chokepoint for AI x bio risk: Synthesis screening is one of the most tractable intervention points for AI-enabled biorisks, as almost every pathway to engineering a pandemic-capable pathogen relies on synthetic nucleic acids. See recent screendna.org open letter signed by IBBIS alongside leaders from OpenAI, Google DeepMind, Anthropic, Microsoft AI and Meta
• Screening tools enable adoption: To reach the 80% threshold of adoption identified by the recent Sentinel field strategy guide for synthesis screening, it needs to be easy for synthesis companies to adopt screening, which in turn makes it tractable to request that screening become mandatory
• Tools need updates to keep pace with emerging threats: Screening must be good enough to actually reduce AI-enabled biorisk; tools need constant development to ensure they keep pace with emerging threats and commercial tools may focus on regulatory compliance instead of flagging more speculative risks

Why fund Commec specifically?

• Increase global adoption of synthesis screening: Commec is one of only two free, actively-maintained synthesis screening tools (alongside SecureDNA) and covers the largest set of global regulations of any screening tool
• Open-source tools support AI risk evaluations: Commec is the only actively-maintained tool with fully open databases and code, which means it's often the first choice for red-teamers and evaluators, including (among users we’ve liaised with) internal teams in frontier labs, biodesign tool developers, and AI safety researchers in both the US and China
• Synthesis companies will thank you: Commec is already used by companies across the US, EU, and China, including large synthesis providers like Dynegene and automated biofoundries like SIAT
• Strong public track record: including participating in the Microsoft-led Paraphrase project (covered in Nature, WaPo, MIT Tech Review) and SBRC standards development
• Reinforces IBBIS's policy engagement and standards development: our technical expertise in screening lends up credibility with policymakers, allows us to engage with key industry stakeholders, and refines our efforts to develop engineering and ISO standards for synthesis screening

What have we shipped recently?

• We released Commec v2.0 in July 2026, which included:
  • 100x reduction in database download size, with 100x speedup in screening
  • Flagging logic that allows users to see where pathogens are regulated; see map of current coverage here
  • Improved sensitivity to AI-redesigned proteins when split into short fragments, see paper with Microsoft Research, “The limits of sequence-based biosecurity screening tools in the age of AI-assisted protein design”
• We are currently testing:
  • "Commec-In-A-Box": distributing Commec on low-cost refurbished Linux hardware, enabling adoption by synthesis companies with limited compute resources (we have sent out our first box and have 5 active requests)
  • A graphical user interface: a beta tester experimenting with the GUI said it was "exactly the kind of solution we wanted to build so that other departments without CLI experience can use it"
• The tool currently screens 200bp sequences in under 1 second, can screen fragments as small as 50bp, and has been measured to have an <2% false positive rate on engineered sequences and <0.5% false positive rate on AI-redesigned sequences, while still catching >97% of AI-redesigned proteins.

How will this funding be used?

In addition to responding to user requests and ongoing development to maintain alignment with best practices, the next year includes:

• Improving resilience to AI threats, including in-code assembly of short fragments (oligo pools) and a collaboration with Fourth Eon Bio and Microsoft to empirically validate the limits of HMM-based detection of AI-designed proteins
• Expanded adoption in China, India, South Korea, South Africa and Brazil (driven by an IBBIS separately-funded adoption campaign, though ongoing technical support is necessary to be responsive to user requests)
• Benchtop integration: since Commec is the only free tool that operates fully offline, we've received significant interest from three benchtop device manufacturers, operating in three continents, and we are currently mentoring 3 ERA fellows focused on benchtop security
• Continued collaboration with policymakers in Canada, the USA, the UK to ensure there is a freely-available screening tool that meets all regulatory requirements for existing or planned synthesis screening mandates
• Improved explanations to assist companies in interpreting screening flags (for example, offering different flags for common viral vector parts vs. rare viral proteins, expanding regulatory coverage to additional regions)

Current funding situation

IBBIS has not been funded to support the next year of the Commec project and without additional funding the project will shift into maintenance mode in November 2026. The project was initially incubated by NTI, and was then supported by the unrestricted grant provided to IBBIS at its founding in 2023. IBBIS has not received any Commec-specific funding.

• Minimal: $206,000. Covers compute costs, our project lead (maintained one full-time developer) and overhead to support IBBIS’s general operations.
• Full: $343,500. Would allow us to complete our 2027 roadmap
• Accelerated: $420,500. Full 2027 roadmap, quicker.

If we receive less than the minimum: This extends the time before we need to transition the project to maintenance mode, but the extension is less than one year.

To note: This funding request has been uploaded on Lightcone Commons.

Who is on your team?

IBBIS, the International Biosecurity and Biosafety Initiative for Science, was founded in 2023 as one of Founders Pledge’s 10 Biggest Bets. IBBIS is a nonprofit Swiss Foundation led by Piers Millett (LinkedIn, former iGEM and Future of Humanity Institute). The Commec project was led by Tessa Alexanian (website, Google Scholar, LinkedIn) from 2024-2026, who is now transitioning to a role at Sentinel Bio. The project will be led going forward by Manu Shivakumara (Google Scholar), previously a Senior Bioinformatics Engineer on the project.

What are the most likely causes and outcomes if this project fails?

• At present, our primary obstacle is funding: we need to keep at least one developer on to support the project.
• A free tool does not automatically mean rapid adoption, and we have hired a Partnerships and Adoption Lead - we have funding through grantmaking.ai to support them for 6 months, and are actively fundraising for additional support.
• We face a few unique challenges for an open-source, fully-local screening tool: it's challenging for us to guarantee that our users will update their software, and we've been reluctant to screen for "non-obvious" sequences, as we want to remain; we still believe Commec is a significant improvement beyond the baseline (or nonexistent) screening we observe at many companies, and have ideas about how to handle this (e.g. federated authorization for downloading detailed annotations, local license keys with expiry dates) but are unlikely to be able to pursue this without