From Scott and Fuller's dog experiments to Russian silver foxes, animal cloning, and CRISPR, scientists have learned a great deal about how genes influence behavior. But can we use that knowledge to produce exceptional working dogs?
OVERVIEW: Genetic engineering and cloning may improve the health and breeding of military working dogs, but research shows that complex behavior also depends on development, experience, training, and the environment.
Genetics can increase the likelihood that a dog will possess desirable working characteristics. Cloning can reproduce an exceptional dog's nuclear genetic makeup, and modern gene-editing technology can change specific biological characteristics. What science has not yet demonstrated is the ability to reliably engineer the complete behavioral abilities of an exceptional military working dog.
A newly announced company called Warhound intends to explore some of these possibilities. Its ambitions raise fascinating questions about animal genetics, behavior, and development.
What Is Warhound Trying to Accomplish?
On October 22, 2026, Warhound announced that it had raised $200 million to develop an American supply of advanced working dogs for military, law-enforcement, and emergency-response applications.
The company, conceived by entrepreneurs Ben Lamm and Palmer Luckey, proposes combining advanced breeding, cloning, genetic technology, training, and specialized equipment.
Its objectives include producing healthier dogs, improving working capabilities, and reducing reliance on imported animals.
These are worthwhile subjects for scientific investigation. Producing suitable working dogs is expensive, time-consuming, and frequently unsuccessful. Not every dog entering a training program eventually qualifies for operational service.
However, there is a distinction between producing a genetically promising puppy and producing an adult dog capable of performing demanding assignments. To understand that distinction, and appreciating the difficulty, we should begin with research conducted more than 60 years ago.
Haven't Dog Breeders Been Doing Something Similar for Generations?
In a sense, dog breeders have been attempting to accomplish something similar for generations through purebred breeding, linebreeding, and, in some cases, close inbreeding. These practices are not cloning in the biological sense, since each puppy inherits a new combination of genetic material from its parents. However, the objective is often similar: to preserve and reproduce desirable inherited characteristics with greater predictability.
Breeders may carefully select dogs from generations of successful working animals, repeatedly concentrating particular bloodlines in an effort to produce consistent physical and behavioral traits. Yet anyone familiar with breeding working dogs knows that even puppies from the same litter can develop remarkably different personalities, motivations, sensitivities, and working abilities. Some become exceptional working dogs, while their littermates may be better suited to entirely different lives. Even prolonged linebreeding does not eliminate this variation.
Genetic recombination, developmental differences, environmental influences, and individual learning histories continue to produce unique animals. Cloning offers a much closer reproduction of an individual dog's nuclear genetic makeup than traditional breeding, but it does not eliminate the developmental processes that make each dog an individual. This is one reason the work of Scott and Fuller remains so relevant to modern genetic engineering.
What Did Scott and Fuller Discover About Dog Genetics and Behavior?
John Paul Scott and John L. Fuller demonstrated that canine behavior is influenced by heredity, but that genetic differences do not determine every aspect of an individual dog's development or performance.
Their landmark book, Genetics and the Social Behavior of the Dog (1965), summarized years of experimental research involving five breeds: Basenjis, Beagles, Cocker Spaniels, Shetland Sheepdogs, and Wire Fox Terriers.
They examined behavioral development, emotional responses, social behavior, learning, and problem-solving abilities. Their research went beyond comparing breeds. They also conducted breeding experiments, studied crosses between breeds, manipulated aspects of the dogs' environments, and investigated the effects of different rearing conditions.
The breeds differed in several behavioral characteristics. For example, Beagles performed particularly well on scent-trailing tasks, while Basenjis excelled in certain manipulation problems.
Yet no breed demonstrated universal superiority across all the problem-solving tasks.
The researchers also observed substantial individual differences within breeds.
These findings are important because they challenge the assumption that an animal possesses a single general quality called intelligence, trainability, or working ability. A dog can excel at one task and perform poorly at another.
Scott and Fuller also showed that emotional characteristics could influence performance. A dog's response to an unfamiliar situation might affect its ability to demonstrate a behavior that it was otherwise capable of performing.
This distinction matters enormously in working-dog training. A dog that fails an exercise is not necessarily incapable of learning or performing it. The animal's motivation, emotional response, previous experience, and immediate circumstances may all influence the outcome.
The lesson is not that genetics are unimportant. Scott and Fuller found substantial hereditary influences. The lesson is that the relationship between genes and behavior is complicated, and environmental conditions can influence how inherited characteristics are expressed.
What Did the Russian Domesticated Fox Experiments Demonstrate?
The Russian silver fox experiments demonstrated that selectively breeding animals for their responses to humans can produce substantial behavioral changes over generations.
Beginning in the Soviet Union, geneticist Dmitry Belyaev and his colleagues selected foxes partly according to their willingness to approach humans and their relative lack of defensive responses. Over successive generations, the selected populations developed pronounced differences in their interactions with people.
The experiments are often presented as evidence that selecting animals for tameness can also produce changes in appearance, physiology, and development. However, there is an important scientific controversy.
A 2020 review by Kathryn Lord and colleagues examined the history of the fox populations and questioned whether some physical characteristics attributed to the experiment were already present in their fur-farm ancestors. The authors also challenged the idea that domestication necessarily produces a universal collection of physical and behavioral characteristics.
This criticism does not eliminate the importance of the behavioral selection results. Rather, it reminds us to distinguish what an experiment demonstrated from broader explanations subsequently attached to it. The fox research shows that inherited behavioral tendencies can respond to sustained selection. It does not establish that selecting for one characteristic will automatically produce every other characteristic we might want.
That distinction is directly relevant to military working dogs. Selecting for stronger pursuit behavior, for example, does not necessarily produce better environmental stability, handler cooperation, or reliable behavioral control. Those characteristics may have different genetic and developmental influences.
Can Scientists Actually Change Animal Behavior by Manipulating Genes?
Yes. Experiments involving several species demonstrate that changing genetic activity can alter behavior, sometimes dramatically.
· In 1998, researchers Mario de Bono and Cornelia Bargmann studied the microscopic roundworm Caenorhabditis elegans. They demonstrated that natural variation in a gene called npr-1 influenced whether worms tended to feed alone or aggregate with other worms. Experimental genetic changes could alter this feeding pattern. This was important because it established a direct connection between a particular genetic mechanism and an observable behavioral difference.
· Research in fruit flies has also identified genes involved in courtship behavior, neural development, and other behavioral processes.
· Experiments in mice have shown that modifying genes associated with neurotransmitter regulation can alter aggression-related responses and other behaviors.
But these experiments also demonstrate an important limitation. A gene rarely operates in complete isolation from the rest of an animal's biology. Changing a gene associated with one behavior may also affect sensory processing, physiological regulation, development, or other behavioral characteristics. Scientists call this pleiotropy, meaning that a single gene can influence multiple traits. A genetic modification that produces a desirable result in one laboratory test may therefore have unintended consequences elsewhere. And a behavioral change demonstrated in a worm or mouse does not establish that the same intervention will produce a useful result in a dog.
The nervous system, developmental processes, and behavioral demands of a military working dog are considerably more complicated.
What Can CRISPR Genetic Engineering Actually Do?
CRISPR-based gene editing allows researchers to make targeted changes to DNA. Its potential is especially promising when scientists understand the biological function of the gene being modified.
For example, researchers have genetically modified pigs to resist porcine reproductive and respiratory syndrome virus, an important disease affecting swine. The virus depends on particular features of a cellular protein called CD163. By modifying the relevant genetic sequence, researchers produced pigs resistant to infection under experimental conditions. This is a meaningful example of successful genetic engineering. The scientists were addressing a relatively well-defined biological mechanism. They knew something important about how the virus entered cells and could test whether the genetic modification interrupted that process.
Other research has demonstrated that seemingly similar edits can produce different results, reinforcing the need to validate the exact modification rather than assume that editing a particular gene guarantees the intended effect.
Now consider the difference between modifying a viral entry mechanism and attempting to produce a dog with exceptional courage, persistence, scent-detection ability, and cooperation with its handler. There is no established single gene for an exceptional military working dog. Complex behavioral characteristics generally involve numerous genes, interacting physiological systems, and developmental influences. CRISPR may eventually contribute to improvements in specific aspects of canine health or performance. But successful editing of DNA is not the same as successfully predicting the adult animal's behavior.
Can Cloning Reproduce an Exceptional Working Dog?
Cloning can reproduce much of an exceptional dog's genetic foundation, but it cannot reproduce the animal's entire developmental history or learned behavior.
Cloning has already been investigated in working dogs.
A study published in Behavior Genetics in 2014 examined seven puppies cloned from an elite drug-detection dog and compared them with four naturally bred control dogs. The researchers reported an 86% selection rate among the cloned dogs, compared with a reported 30% among naturally bred dogs. This is encouraging evidence for cloning technology. However, the study involved very small groups, and its findings cannot establish that cloning will produce the same results across different donors, working assignments, and training programs.
We also need to distinguish qualification for service from long-term operational performance. A dog that successfully completes a detection-dog training program may not necessarily demonstrate the same working longevity, health, or reliability as the animal from which it was cloned.
Cloning has also been used in horses, including animals valued for their athletic performance and breeding potential. But cloning a successful horse does not transfer the original animal's conditioning, training, experiences, or learned responses. The same applies to dogs.
A cloned dog does not inherit the original dog's memories. It must still develop, learn, mature, and interact with its environment.
Why Doesn't Identical DNA Guarantee Identical Behavior?
Identical or nearly identical nuclear DNA does not guarantee identical behavior because genes operate within developing biological systems that are influenced by environmental conditions.
Consider two puppies with essentially the same nuclear genetic makeup. They may develop in different surrogate mothers, experience different prenatal conditions, receive different maternal care, and encounter different environments after birth. Their experiences with people, other dogs, unfamiliar surfaces, noises, objects, and training may differ substantially. Those experiences can influence later behavior.
Epigenetics adds another dimension. Epigenetic processes regulate gene activity without necessarily changing the DNA sequence itself. Some epigenetic mechanisms are involved in normal development, while others may respond to environmental conditions. However, we should avoid assuming that every behavioral difference between genetically similar animals results from epigenetics.
Prenatal development, learning, random developmental variation, health, and numerous other processes can contribute. The specific role of persistent epigenetic changes in adult working-dog performance remains an area requiring additional research.
Scott and Fuller's findings are relevant here because they demonstrated that hereditary influences and developmental conditions must be considered together.
Can Genetic Testing Predict a Dog's Future Behavior?
Genetic information can help predict some behavioral tendencies, particularly at the population level, but predicting the behavior of an individual dog remains much more difficult.
A major study published in Science in 2022 examined behavioral information from more than 18,000 dogs and genetic information from more than 2,000. The researchers found that many behavioral characteristics were heritable. However, breed explained approximately 9% of individual behavioral variation in the population studied. This figure does not mean that genetics explains only 9% of behavior. It means that breed identity, by itself, was a limited predictor of an individual dog's behavior in that study.
We should also be careful about applying results from a broad population of companion dogs directly to highly selected working-dog lines. Purpose-bred populations may differ substantially in their genetic characteristics and behavioral distributions.
Research into working-dog breeding has shown that genomic information, pedigree analysis, and estimated breeding values may improve selection decisions. The important distinction is between improving the probability of producing desirable traits and guaranteeing that a particular puppy will possess them. The former is scientifically supported. The latter remains much more difficult.
Can Artificial Intelligence Calculate the Genetics of a Perfect Working Dog?
Artificial intelligence can analyze enormous quantities of genetic information, identify patterns that would be difficult for human researchers to recognize, and help breeders predict which combinations of dogs are more likely to produce desirable characteristics. However, having the mathematical ability to analyze millions of genetic variables is not the same as knowing precisely what those variables will produce in a living animal.
Recent research involving guide dogs demonstrates both the promise and limitations of this approach. A 2026 study of Labrador Retrievers found that incorporating genomic information improved predictions of inherited behavioral characteristics compared with pedigree information alone. Nevertheless, the predictions remained far from perfect. Another study involving more than 3,000 dogs found that previously proposed individual genetic markers were not reliable predictors of the behavioral traits examined.
We also need to consider the familiar bell curve. When breeders select animals for desirable characteristics, they may shift the average performance of future generations. More puppies may possess the characteristics necessary for demanding work. But improving the average does not eliminate individual variation. Puppies from carefully selected working lines still differ in motivation, environmental responses, learning, health, and eventual working ability.
Cloning changes part of that equation because cloned animals possess essentially the same nuclear genetic makeup. Yet research involving cloned dogs has documented differences among animals produced from the same donor. Other experiments have found greater behavioral consistency among clones than among control dogs. Both findings matter. Cloning may reduce some variation without eliminating it.
CRISPR technology introduces another possibility by allowing scientists to modify particular genetic sequences. This may be especially useful when a desirable outcome depends on a well-understood genetic mechanism. However, complex behavioral characteristics are influenced by numerous genes and their interactions with development and experience. Editing one or several genes does not establish that the resulting animal will possess the complete combination of abilities required for operational work.
This brings us back to Scott and Fuller. Their experiments demonstrated that inherited behavioral characteristics are important, but they also showed that environmental conditions and individual development influence how those characteristics are expressed.
AI may eventually help produce more healthy, capable working dogs and reduce the number that fail to qualify for service. That would be a meaningful achievement. But even an exceptionally sophisticated mathematical model cannot guarantee an outcome when many of the biological relationships it must calculate remain incompletely understood.
The realistic scientific objective is not necessarily to eliminate the bell curve. It is to improve the probability of producing dogs that meet demanding standards while preserving health, welfare, and genetic diversity.
Why Is Producing an Exceptional Working Dog So Complicated?
An exceptional working dog must possess more than one desirable characteristic, and those characteristics must function together.
Working-dog trainers often discuss prey drive, defense drive, pack drive, hunt behavior, persistence, environmental confidence, and handler responsiveness. These terms describe useful aspects of behavior, but they should not automatically be treated as separate genetic units.
Consider prey drive. In working-dog terminology, this can include orienting toward moving objects, pursuing, grabbing, maintaining possession, and engaging in related activities.
Those behaviors may involve different motivational, sensory, neurological, and learned processes. Likewise, what trainers describe as defense drive may involve several forms of defensive responding that differ according to context. Pack drive is another traditional working-dog term, often used to describe aspects of social affiliation and cooperation. It should not be confused with a single scientifically established genetic mechanism.
A dog may possess intense motivation yet lack the behavioral flexibility needed to perform reliably. Another may show excellent environmental confidence but insufficient persistence for a demanding search assignment. A third may perform exceptionally well with one handler but respond differently when the handler, training conditions, or environment changes.
From my perspective as a dog trainer, the distinction is between possessing a desirable characteristic and being able to use it appropriately. High motivation is not the same thing as reliable performance. Nor is suppressing a dog's behavior necessarily evidence that the animal has learned to regulate its responses effectively.
A successful working dog must be able to engage in demanding activity while remaining responsive to relevant cues and changing circumstances. That combination cannot be reduced to a simple genetic label.
Could Better Genetics Still Improve Military Working Dogs?
Absolutely. Genetic technologies may offer substantial improvements without making behavior completely predictable.
One promising application is reducing the incidence of particular inherited diseases. Another is improving breeding decisions through genetic information combined with health records, pedigree analysis, and carefully measured behavioral performance. Cloning may also help preserve valuable genetic material from exceptional animals. These approaches could potentially improve the percentage of dogs that successfully complete training and remain healthy throughout their working lives. However, improvements should be evaluated against the complete breeding and training process.
Cloning, for example, involves reproductive procedures that can carry costs and animal-welfare concerns. A higher qualification rate among live-born cloned puppies would not, by itself, establish that the entire process is more efficient or beneficial. There are also questions about maintaining genetic diversity. Repeatedly reproducing a small number of exceptional genotypes could have advantages, but responsible breeding programs must consider long-term population health and the consequences of concentrating particular genetic characteristics.
The objective should not simply be producing more dogs that pass an initial test. It should be producing healthy animals capable of performing their intended assignments reliably and sustainably.
How Would We Know Whether Warhound's Approach Works?
The strongest evidence would come from independently evaluated outcomes across the complete lives of the dogs produced.
A meaningful evaluation would examine the proportion of dogs successfully qualifying for service, the reasons other dogs were disqualified, veterinary health, operational performance, working longevity, and animal welfare. It would also need to account for differences in breeding, early development, training, handlers, and assignment requirements.
Suppose a new breeding program achieves a substantially higher qualification rate than a conventional program. That would be encouraging. But the improvement might reflect better genetic selection, better veterinary care, improved developmental practices, more effective training, or some combination of these factors.
Without appropriate comparisons, we could not determine how much improvement came from genetic engineering itself.
This is why research design matters. The question is not simply whether a program produces exceptional dogs. It is whether the program produces better outcomes than reasonable alternatives, and whether those improvements can be reproduced consistently.
What Should We Conclude About the Future of Genetically Engineered Working Dogs?
Genetic engineering may improve the production of military working dogs, but current science does not establish that we can reliably design the complete behavioral profile of an exceptional operational animal.
The history of behavioral genetics provides good reasons for both optimism and caution. Scott and Fuller demonstrated that inherited differences matter, but also that environmental conditions and emotional characteristics influence performance. The Russian fox experiments showed that sustained selection can change behavioral tendencies across generations. Research involving worms, mice, livestock, and other animals has demonstrated that genetic manipulation can produce measurable biological and behavioral changes.
Cloning research has shown that valuable genotypes can be reproduced, sometimes with encouraging working-dog selection outcomes. And CRISPR has opened possibilities that earlier generations of geneticists could scarcely have imagined. But none of these findings establishes that we can assemble a collection of desirable genetic characteristics and reliably produce a fully trained, operationally successful dog.
Warhound's project may contribute important advances to canine genetics, veterinary medicine, breeding, and working-dog development. Its greatest contribution may ultimately be improving the probability of producing healthy, capable working dogs rather than making their individual performance completely predictable. For me, that is the most interesting possibility.
The science suggests that we should continue improving our understanding of genetics while giving equal attention to developmental biology, learning, motivation, health, and the environments in which dogs live and work.
We may become much better at producing the genetic potential for an exceptional working dog. But producing the exceptional working dog itself remains a much larger biological and developmental undertaking. I am very interested in how these experiments turn out.
Glossary
Behavioral genetics: The study of how genetic variation contributes to differences in behavior.
Cloning: Producing an organism with essentially the same nuclear genetic sequence as another organism, commonly through somatic cell nuclear transfer.
CRISPR: A family of technologies used to make targeted changes to genetic material.
Epigenetics: Processes that regulate gene activity without necessarily altering the underlying DNA sequence.
Genotype: An organism's genetic constitution, either generally or at specified genetic locations.
Phenotype: The observable characteristics of an organism, resulting from genetic, developmental, environmental, and other influences.
Heritability: A population-specific estimate of the proportion of variation in a measured trait statistically attributable to genetic differences under the conditions studied.
Polygenic trait: A characteristic influenced by numerous genetic variants.
Pleiotropy: The influence of a gene on more than one characteristic.
Genomic selection: Using genetic information to help predict breeding value and improve selection decisions.
Gene-environment interaction: A situation in which the effects of genetic differences depend on environmental conditions, or environmental effects depend on genotype.
Operational qualification: Meeting the standards established for a particular working assignment. Qualification does not necessarily establish long-term operational success.
Scholarly Bibliography
Bray, E. E., Otto, C. M., Udell, M. A. R., Hall, N. J., Johnston, A. M., & MacLean, E. L. (2021). Enhancing the selection and performance of working dogs. Frontiers in Veterinary Science, 8, 644431. https://doi.org/10.3389/fvets.2021.644431
Chen, F. L., Zimmermann, M., Hekman, J. P., Lord, K. A., Logan, B., Russenberger, J., Leighton, E. A., & Karlsson, E. K. (2021). Advancing genetic selection and behavioral genomics of working dogs through collaborative science. Frontiers in Veterinary Science, 8, 662429. https://doi.org/10.3389/fvets.2021.662429
Choi, J., Lee, J. H., Oh, H. J., Kim, M. J., Kim, G. A., Park, E. J., Jo, Y. K., Lee, S. I., Hong, D. G., & Lee, B. C. (2014). Behavioral analysis of cloned puppies derived from an elite drug-detection dog. Behavior Genetics, 44(1), 68–76. https://doi.org/10.1007/s10519-013-9620-z
de Bono, M., & Bargmann, C. I. (1998). Natural variation in a neuropeptide Y receptor homolog modifies social behavior and food response in C. elegans. Cell, 94(5), 679–689. https://doi.org/10.1016/S0092-8674(00)81609-8
Lord, K. A., Larson, G., Coppinger, R. P., & Karlsson, E. K. (2020). The history of farm foxes undermines the animal domestication syndrome. Trends in Ecology & Evolution, 35(2), 125–136. https://doi.org/10.1016/j.tree.2019.10.011
Morrill, K., et al. (2022). Ancestry-inclusive dog genomics challenges popular breed stereotypes. Science, 376(6592), eabk0639. https://doi.org/10.1126/science.abk0639
Scott, J. P., & Fuller, J. L. (1965). Genetics and the social behavior of the dog. University of Chicago Press.
Whitworth, K. M., et al. (2016). Gene-edited pigs are protected from porcine reproductive and respiratory syndrome virus. Nature Biotechnology, 34, 20–22. https://doi.org/10.1038/nbt.3434
Company announcement and background reporting
Warhound. (2026, October 22). Warhound launches with $200 million to rebuild America's military working dog supply, using advanced genetics and defense technology. PR Newswire. https://www.prnewswire.com/news-releases/warhound-launches-with-200-million-to-rebuild-americas-military-working-dog-supply-using-advanced-genetics-and-defense-technology-302917498.html
