Tech giants warn AI safety gaps could hand bioweapons to bad actors

The CEOs of some of the biggest artificial intelligence (AI) companies, including OpenAI, Anthropic, Microsoft AI and Google DeepMind have teamed up with biotechnology experts to urge the United States to pass laws requiring safety screenings of synthetic DNA purchases.

Synthetic DNA refers to artificially manufactured genetic material, which can be custom-made by companies and ordered online, and delivered like any other laboratory supply.

The letter is directed at the US Congress, but the issue has global implications as the material can be ordered and shipped worldwide.

“The ability to order synthetic DNA online has accelerated vaccine development, powered basic research, and made it possible for small teams to access capabilities that used to be confined to major institutions,” the signatories wrote.

But it also creates a risk, as, in theory, a bad actor could order DNA sequences designed to recreate dangerous pathogens.

Synthesized DNA has a wide range of uses. It helps create life-saving medicines, engineer microorganisms and store vast amounts of digital data.

Some companies already check orders and customers voluntarily, but there is no current legal requirement to do so.

The letter noted that screening “is also one of the best understood and least disruptive biosecurity measures available”.

The signatories also ask for mandatory record-keeping in case there is a need to track suspicious activity, and not only after the fact, but because awareness of traceability deters misuse, according to the letter.

“So that any threat that might evade initial screening can be traced back to its source — including when individual sequences would not raise concern in isolation,” the letter read.

The authors noted that while this issue is not new, the pace of progress in artificial intelligence is and AI systems can now outperform PhD-level virologists on questions about highly technical laboratory procedures.

As systems rapidly improve, experts warn that the knowledge barriers that have historically prevented bad actors from obtaining biological weapons will “meaningfully erode”.

Earlier this year, researchers worldwide also warned that unrestricted access to certain biological datasets could enable AI systems to help design or enhance dangerous viruses, calling for stronger safeguards to prevent misuse.

In an open letter published in February, researchers from leading institutions, including Johns Hopkins University, the University of Oxford and Stanford University, argue that while open-access scientific data has accelerated discovery, a small subset of new biological data poses biosecurity risks if misused.

To guarantee safe access, the letter called for specific technical tools that would enable data providers to verify legitimate users and track misuse, such as watermarking and audit logs.

What is the situation in Europe?

The European Commission published the EU Biotech Act in 2025, a proposal to regulate the EU’s biotechnology and biomanufacturing sectors, where it warns that “biotechnology introduces new biosecurity risks as the wider accessibility of these technologies increases their potential for misuse, posing significant health threats”.

There is currently no European law to regulate the purchases of these materials, and the Commission noted that divergent or absent national rules fail to offer a level playing field to competitors and weaken prevention.

The Act identifies synthetic nucleic acid sequences (DNA and RNA) as “biotechnology products of concern,” and proposes a new EU-wide framework.

The Commission’s proposal introduces some of the same ideas as the US letter, including customer verification and reporting of suspicious orders.