A Breakthrough That Could Reshape Biology

In March 2026, the scientific world witnessed a milestone that some researchers compare to the "ChatGPT moment" for synthetic genomics. The Arc Institute has unveiled Evo2, an artificial intelligence model that doesn't just read and interpret DNA sequences — it can now write them from scratch.

"It's cool, but it's not there yet," says Nico Claassens, synthetic biologist at Wageningen University in the Netherlands. "The need to synthesize and test AI-generated genomes at scale is one major hurdle. Another is designing genomes that can direct all the essential functions of even the simplest life forms, let alone those of more-complex cells."

From Bacteria to Humans

The Evo2 model, detailed in a Nature paper published on 4 March, was trained on trillions of base pairs of DNA — sequences from bacteria, archaea, and eukaryotes. Unlike its predecessor, Evo2 has learned to recognize complex genomic features like regulatory DNA, splice sites, and protein-coding sequences that humans struggle to identify.

"These AI models are the 'ChatGPT moment' for synthetic genomics," says Patrick Yizhi Cai, genome engineer at the University of Manchester. "You can start writing things that never existed in nature."

A New Era for Synthetic Life?

Historically, creating synthetic genomes meant editing existing ones — like removing commas from a book or recoding bacteria at specific locations. Evo2 changes that paradigm. In 2025, researchers used early Evo models to design viral genomes that successfully infected Escherichia coli. Now, the same team, led by computational biologist Brian Hie, has created a genome inspired by Mycoplasma genitalium — a bacterium whose 580,000-nucleotide genome was the first ever chemically synthesized in 2008.

"These AI models are the 'ChatGPT moment' for synthetic genomics," says genome engineer Patrick Yizhi Cai at the University of Manchester. "You can start writing things that never existed in nature."

The Path Forward

Scientists remain cautious. Creating functional genomes requires more than just design — it demands large-scale synthesis and biological testing. Yet, the field is optimistic. "Once we can write genomes that work, the implications for medicine, agriculture, and synthetic biology could be revolutionary," says one researcher.

As Evo2 proves that AI can now generate novel genetic sequences, humanity stands on the brink of a new frontier — where the code of life itself can be written by algorithms rather than evolved by nature.

The Implications

If successful, this technology could enable:

  • Customized therapies designed from genetic blueprints that never existed in nature
  • Optimized organisms engineered for specific functions in agriculture and medicine
  • New life forms with novel metabolic pathways and genetic architectures

The question remains: how far can AI take us in redefining life itself?