Colossal Biosciences Leverages AI to Advance the De-Extinction of the Dodo

Since its inception in 2021, the healthtech and genetics firm Colossal Biosciences has set its sights on one of the most ambitious frontiers in modern science: the de-extinction of species lost to time. Among its primary targets is the Dodo, the flightless, iconic bird endemic to Mauritius. When Portuguese sailors first encountered the species in the 16th century, they documented a creature that had evolved in total isolation. However, the Dodo’s narrative turned tragic shortly thereafter. The arrival of invasive species and intensive human hunting proved catastrophic, leading to the bird’s total disappearance within a relatively short span of time. For centuries, "going the way of the Dodo" has served as a grim linguistic shorthand for obsolescence and extinction. Now, Colossal Biosciences aims to rewrite that history.

The company, which has already garnered significant attention for its claims regarding the resurrection of the dire wolf, is now applying its advanced genetic toolkit to the Dodo. The challenge is immense. Unlike conventional cloning, which typically requires a complete, viable cell, the company’s strategy relies on "proxy species" engineering. By editing the genomes of living relatives—specifically, types of pigeons—and integrating preserved DNA from the extinct animal, researchers hope to reconstruct a functional version of the lost species.

The Dodo presents a unique set of hurdles. The company currently has access to only a single reliable genomic example, limiting the amount of historical data available to guide their efforts. Furthermore, while the company has successfully applied its technology to mammals like the dire wolf—which could theoretically be birthed by a surrogate—the Dodo poses an avian-specific challenge: the species must be gestated within an egg. This requires an entirely different approach to reproductive biology and embryonic development.

This week, however, Colossal Biosciences announced a significant milestone in this endeavor. The company revealed that it has developed and deployed a sophisticated artificial intelligence tool, dubbed "GenoMorph," to navigate the complexities of the Dodo’s genetic map. This tool represents a critical shift in how geneticists approach the "de-extinction" problem. By training the AI on the DNA of various extant species, the researchers have enabled GenoMorph to predict which sections of a genome are associated with specific biological traits, such as beak morphology, feather structure, or skeletal composition.

The utility of GenoMorph lies in its ability to parse through the fragmented Dodo DNA currently available. By identifying which specific regions of the existing Dodo genome are responsible for the physical characteristics that define the bird, the team has successfully narrowed their focus. As of this latest update, the scientists have identified 2,300 specific DNA targets that are critical for achieving a "Dodo-ish" phenotype. Furthermore, the company reported that it has successfully utilized lipid nanoparticles to conduct genomic editing within an egg for the first time, a development that addresses one of the most significant technical barriers in avian genetic engineering.

It is vital to distinguish between true biological replication and the functional approach Colossal Biosciences is taking. The company is not attempting to recreate the original Dodo down to every single base pair of its ancient genetic code. Instead, the strategy is to create a "functional" Dodo. This involves taking the genomic sequence of a living, closely related pigeon and systematically filling in or replacing segments of that sequence with the 2,300 identified Dodo DNA targets. The goal is to produce an organism that, while genetically distinct from the original in some respects, expresses the physical and behavioral traits that made the Dodo unique.

This distinction remains at the heart of an ongoing, vigorous debate within the scientific community. Whether or not this process constitutes the "true" return of a species is a question that challenges our fundamental definition of what a species is. Critics have frequently objected to the framing used by Colossal Biosciences, arguing that the term "resurrection" is scientifically misleading when applied to proxy-based engineering. These critics have voiced similar concerns regarding the company’s other high-profile projects, which include efforts to bring back the woolly mammoth, the thylacine, and the giant Moa.

For the leadership at Colossal Biosciences, however, these critiques miss the larger point about the fluidity of biological classification. Beth Shapiro, the company’s Chief Science Officer, argues that the rigid boundaries we often place around species are more philosophical than they are biological. "It’s a species definition that’s based on an arbitrary threshold of sequence similarity," Shapiro says. "But that threshold is arbitrary. It’s going to be arbitrary no matter who you are."

Shapiro’s perspective suggests that nature itself rarely respects the clean lines humans draw between one species and another. Evolution is a process of constant flux, and the genetic commonalities shared between a pigeon and a Dodo are, in her view, a testament to that shared ancestry rather than an insurmountable wall. By framing the project in this way, Colossal Biosciences is attempting to move the conversation away from the binary of "extinct vs. alive" and toward a more nuanced understanding of "functional restoration."

The role of artificial intelligence in this process cannot be overstated. Without tools like GenoMorph, the task of identifying which of the thousands of potential genetic sequences contribute to the Dodo’s specific traits would be a process of trial and error that could span decades. By accelerating the identification of these traits, the AI allows the team to prioritize their editing efforts on the most meaningful parts of the genome. It is a marriage of high-compute power and wet-lab biology that defines the modern approach to synthetic biology.

As the company moves forward, the scientific and ethical scrutiny will likely only intensify. The prospect of bringing back extinct animals touches on deep questions regarding our responsibility to the ecosystems we have altered and the ethics of manipulating life in a laboratory setting. Colossal CEO Ben Lamm and his team remain focused on the technical milestones, viewing their work as a necessary evolution in biotechnology.

In recent discussions regarding these developments, Lamm and Shapiro emphasized that the updates are not merely about the Dodo, but about the refinement of a platform. The ability to use lipid nanoparticles for in-egg editing, combined with the predictive power of AI, creates a pipeline that could potentially be applied to a variety of avian conservation and restoration projects. While the Dodo remains the most prominent face of their efforts, the underlying technologies are being designed to handle the complexity of complex vertebrate genomes across the board.

For now, the Dodo remains a symbol of a lost past, but the work being conducted in Colossal’s labs suggests that the future of extinction may not be as permanent as once thought. The team continues to refine its list of genetic targets and improve its methods of embryonic integration, inching closer to a day when the "functional" Dodo may finally emerge. Whether the result is accepted by the broader scientific community as a true resurrection or as a new, engineered entity remains to be seen. What is clear is that the pursuit of the Dodo has moved from the realm of historical curiosity into the cutting-edge of genetic engineering, where the line between history and the future is being rewritten one sequence at a time.