Artificial biological intelligence
Artificial biological intelligence (ABI) is a term for artificial intelligence systems that can read, design, construct, and activate ("boot up") genomes and, ultimately, living organisms. It was introduced by the British physician and molecular biologist Adrian Woolfson, who describes ABI as a catch-all for technologies that treat DNA as a programmable, predictive engineering material rather than something to be studied only after the fact.[1][2] The concept sits at the convergence of synthetic biology, genomics, and modern machine learning, and is the central idea of Woolfson's 2026 book On the Future of Species: Authoring Life by Means of Artificial Biological Intelligence.[3]
Definition and scope
Woolfson uses "artificial biological intelligence" as an umbrella term for systems that can "design, construct, and ultimately 'boot up' living things."[1] The framing draws a deliberate analogy to large language models: where a text model manipulates the letters of a written language, an ABI system manipulates the four-letter alphabet of DNA in order to recognise what a sequence does and to generate new sequences that perform a desired biological function.[1]
In this usage, "intelligence" refers to the capability of the design-and-construction system, not to any awareness on the part of the organisms produced. ABI is therefore distinct from, though related to, broader discussions of intelligence in biological systems.
Core competencies
In a 2025 paper in the journal Molecular Therapy, Woolfson described ABI in terms of four core competencies:[2]
- Predictive – determining what a previously unseen genome encodes, for example which species it specifies or which cell types could be derived from it.
- Generative – proposing candidate genome sequences that would yield a desired anatomy, physiology, or biochemistry, or recognising when a desired outcome is not achievable with natural biological materials.
- Constructive – physically synthesising and assembling the designed genome sequences.
- Instantiative – activating, or "booting up," a genome so that it gives rise to a functioning synthetic organism.
Woolfson argues that combining design and construction is what turns a biologist into an engineer, and that the ability to then instantiate a genome would represent "full mastery of biology as a technology."[1]
Enabling technologies
Woolfson links the feasibility of ABI to several converging advances:
- Genome language models – AI models trained on DNA, such as Evo 2, which uses a context window of roughly one million base pairs so that it can capture interactions between distant regions of a genome.[1]
- Large-scale DNA synthesis – methods that write DNA rapidly, cheaply, and in parallel, which Woolfson identifies as essential to using DNA as an engineering material.[1]
- Genome transplantation – earlier work by Craig Venter showing that a cell can function like a "nanocomputer" running a genome as its "software," illustrating the still-difficult step of getting a synthetic genome to operate inside a cell.[1]
Origin of the term
The term is associated with Adrian Woolfson, co-founder of Genyro, a California biotechnology company specialising in synthetic genome design and construction. Woolfson studied medicine at Balliol College, Oxford, was formerly the Charles and Katherine Darwin Research Fellow at Darwin College, Cambridge, and worked at the MRC Laboratory of Molecular Biology. He is also the author of Life Without Genes and An Intelligent Person's Guide to Genetics.[3] He developed the idea at length in On the Future of Species (MIT Press, 2026).[3][4]
Reception
On the Future of Species, the book in which the concept is elaborated, was reviewed in Science[4] and Kirkus Reviews.[5] Coverage has generally treated ABI as an ambitious, forward-looking framework for the merger of AI and synthetic biology, while noting that many of its capabilities remain aspirational.[1][4]
Risks and ethics
Woolfson emphasises that the same capabilities that make ABI powerful also carry risk. He notes that engineered organisms released into poorly understood ecosystems could cause unintended harm, and that the underlying technologies are inherently dual-use. He argues that they must be developed and deployed "safely, responsibly, ethically, transparently, and equitably."[1][3]
See also
- Synthetic biology
- Synthetic genomics
- Genome editing
- Artificial life
References
- ↑ 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 Strickland, Eliza (29 April 2026). "Can Biologists Rewrite the Genome's Spaghetti Code?". IEEE Spectrum.
- ↑ 2.0 2.1 Woolfson, Adrian (2025). "ABI and generative biology: a new paradigm for gene therapy, genome engineering, and engineered cell therapy". Molecular Therapy. PMC 12126812 Check
|pmc=value (help). PMID 40120587 Check|pmid=value (help). - ↑ 3.0 3.1 3.2 3.3 Woolfson, Adrian (2026). On the Future of Species: Authoring Life by Means of Artificial Biological Intelligence. MIT Press. ISBN 978-0-262-05489-8. Search this book on
- ↑ 4.0 4.1 4.2 "The scope of biological possibility expands". Science. 2026. doi:10.1126/science.aee7892.
- ↑ "On the Future of Species". Kirkus Reviews.
This article "Artificial biological intelligence" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:Artificial biological intelligence. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.
