DNA supply chain
The DNA supply chain is the path from an ordered genetic blueprint to the finished piece of DNA a lab receives in the mail. It is considered the most important control point for preventing AI-designed dangerous genetic material from actually being manufactured.
DNA is the molecule in which living things store their construction plans. Today, such molecules can also be manufactured artificially, letter by letter, at specialized companies. To do this, researchers simply send a text file with the desired sequence of letters to such a company and receive a vial with the finished molecule delivered a few days later. The DNA supply chain is the entire chain of steps this process goes through: ordering, screening of the order, manufacturing, shipping, and use in the lab. The term almost always appears in the news in connection with safety. Because anyone who designs a dangerous blueprint on a computer needs this chain to turn it into something real.
The bottleneck between blueprint and pathogen
A sequence of letters on a screen is harmless. It only becomes dangerous once someone turns it into an actual molecule. This is exactly where the synthesis companies come in, and there are only a few dozen large ones worldwide. From a safety perspective, this narrowness is a stroke of luck: a few points are easier to control than millions of users.
The topic is newly being discussed because of AI models for biology. They can design proteins and genetic sequences that do not occur in nature. This is useful for medicines and vaccines. But it also means that expertise which once required years of study is now partly embedded in a program.
That is why many safety researchers argue as follows: one will never be able to prevent a dangerous sequence from being designed somewhere. But one can prevent it from being manufactured and shipped. The supply chain is thus the last practical hurdle. In the US, there have been requirements for several years that obligate companies receiving government funding to carry out checks.
How synthesis companies screen orders
The core of the control consists of two checks. First, the ordered sequence is compared against databases of known hazardous substances, such as genetic segments from smallpox viruses or toxins like ricin. Second, the customer is checked: Do they belong to a legitimate research institution? Does the order match their field of work? Both checks together are called sequence and customer screening.
This sounds simpler than it is. Dangerous sequences can be modified without losing their effect. The comparison with a database then no longer works reliably. Newer screening methods therefore look not at the letters but at the predicted function of the molecule. AI is also being used for this defense, not just for the attack.
A second problem is workarounds. Anyone who splits a long gene into many short pieces and orders them from different providers in different countries will not stand out to any single one of them. The pieces can later be assembled in one’s own lab. In addition, there are now benchtop devices that let labs produce DNA themselves. This bypasses the supply chain entirely, which weakens its role as a control point.
Why the topic appears in AI news
When major AI labs introduce their new models, they often publish a safety report. This regularly includes a section on biological risks. The supply chain is then usually mentioned as an additional layer of protection: even if a model reveals too much, the order placed with the synthesis company is supposed to fail. Anyone reading these reports can only understand such sentences with this term in mind.
The topic is also present in politics. There are international alliances of synthesis companies that voluntarily commit to screening. Governments and experts are discussing whether such voluntary commitments are sufficient or whether binding laws are needed. This is relevant for investors, because contract research organizations and biotech suppliers would be directly affected.
A common misconception is that the DNA supply chain is a purely logistical matter, like the transport of components. What is almost always meant is the safety issue, not the shipping route. The term should also be distinguished from the CRISPR gene-editing tool: CRISPR modifies existing genetic material, whereas the supply chain delivers new genetic material from scratch.