Some people seem to be born animal whisperers. Whether it’s Steve Irwin, Jane Goodall (RIP to both) or even just your local dog trainer, these people have a certain way with animals. It’s almost as if they can understand them. But what if there was a way to take that further? That’s something pet psychics and animal communication scientists have, through different means, attempted to crack for decades. The Dr Dolittle archetype has been woven throughout human history, promising interspecies communication and a greater connection with a multitude of species. Now, after years of research using machine learning to detect patterns in animal vocalisations, scientists seem closer than ever to achieving interspecies communication. But does that mean we should?
The animal behaviourists using AI in their research are already navigating how to use the rapidly advancing technology to decode animal languages “responsibly” – but, of course, that’s only according to human ethics. Considering humans’ long, long history of exploiting animals for food, materials, medicine and entertainment, it’s hard to imagine a world where animals are consulted on the matter. The field so far has involved a lot of attempted eavesdropping, like the teams collecting squeaks from African striped mice using the same system that underpins large language models such as ChatGPT. Or Project CETI, a nonprofit using machine learning and robotics to listen to and translate the communication of sperm whales.
Along with the African striped mice research, there are three other research projects up for this year’s Dolittle Prize, which will award $100,000 to the team that has made the most significant advances in deciphering animal communication. Sponsored by British billionaire businessman Jeremy Coller, the Dolittle Prize even offers a $10 million investment (or $500,000 cash) to researchers who can communicate with an animal without it realising its communicating with a human. So, what do billionaires really want to be able to say to animals, with the help of AI, if not apologise for exploiting the world for their own gain?
Gašper Beguš, Linguistics Lead at Project CETI, says communicating back to the whales at this point would be “ethically acceptable”. There’s even a PEPP Framework (Prepare, Engage, Prevent, and Protect) ethical framework established around it, which was released in 2025 and which CETI were the first to sign. “Before you understand the language, you cannot talk to them because you cannot minimise the risks [of miscommunication]until you know what you’re actually doing,” Beguš says. “So we’re careful about just listening and passively following them without intervention or playback.” Beguš points to the long process of decoding Egyptian hieroglyphs as a role model for how we might eventually learn to communicate with animals, emphasising the importance of carefully observing patterns before attempting to translate them.
“In reality, tomorrow we will not be able to Google translate talk with other species. There’s a very live question and debate in the field whether that would ever be possible because humans and non-human animals have different referential meanings”
Emma Crowe, a co-author of the PEPP Framework and global waters lead for the MOTH (More-Than-Human Life) Program, says it’s the first framework of its kind. “There are no existing regulations or guardrails to help guide these technologies,” she says. Crowe says the response has been mostly positive. “Those we have talked to are super excited about the idea of shared consensus guardrails and are, in many cases, the ones championing the need for them.” Although the lack of guardrails in AI is hardly unique to animal communication, the rapidly growing technology has largely been given free rein, especially by the Trump administration. This includes weakening the environmental safeguards around running data centres – the likes of which will impact the very animals we wish to connect with, along with the health risks to humans.
With all the talk of AI taking over human jobs, it seems that even the work of a spiritual animal communicator might not be safe. I reached out to Laura Stinchfield, The Pet Psychic, who already runs a business helping people speak to their animals, to ask what she makes of AI’s attempts to do the same. “I don’t believe that AI can replicate real human relationships,” she says. “It’s not as simple as translating some words – you have to connect to their emotions”. Funnily enough, researchers are still scratching their heads on how to do exactly that.
Outside the industry, AI animal communication is often pictured as a Google Translate for different species. But Crowe says there’s a nuance missing there; people studying in this space are having far different conversations. “In reality, tomorrow we will not be able to Google translate talk with other species,” she says. “There’s a very live question and debate in the field whether that would ever be possible because humans and non-human animals have different referential meanings.” In other words: humans and animals don’t share the same systems of reference. Because of this, Crowe sees the Dr Dolittle scenario as not an immediate risk but a situation people are preparing for in the possible future. But that doesn’t mean researchers aren’t making AI-powered leaps and bounds.
In 2025, Project CETI discovered that sperm whale clicks contain vowel-like acoustic patterns. It’s a whole new dimension of whale communication, similar to human speech, that they didn’t know existed before. “Finding language-like features means we have a window into their inner lives and that gives us a unique opportunity to connect with another beautiful species,” Beguš says. “One of the coolest things about it is that what we’re finding has legal implications – if we’re showing that their language is more complex than we used to believe, then their inner lives are probably more complex than we used to believe, and they should have more rights.”
The vision of humans gaining more of an understanding of how animals communicate, and therefore drastically changing our laws to protect animals, is an appealing one. It’s something I personally would love to believe is possible. But there’s no denying that, no matter what communication patterns we find, it will be largely hindered by the fact that we live in an anthropocentric culture that prioritises humans as the central, superior entity on Earth. We have known for decades that pigs, for example, are as intelligent as three-year-old children, but many people still enjoy eating bacon. Even if we could get a clear understanding of what animals are saying to us, the very fact that they are not humans has always been enough to justify their mistreatment.
“The law has always depended partly on what we know about animals. If technology radically changes what we can know about them, it may eventually change what the law believes we owe them”
Still, Kamil Mamak, author of AI Animal Communication: A Potential Source of Moral Revolution and an associate professor of law and philosopher at Jagiellonian University in Kraków, Poland, is hopeful that AI could fundamentally change what we know about animals and how we treat them. “Communication can make animals appear more human-like and relatable to us; this can increase empathy,” he says. “As a result, we might care more about animals both because we know more about what they experience and because communication makes it psychologically harder to see them as radically different from us.” This, says Mamak, could carry over into welfare law, where better evidence about pain and psychological suffering could influence regulations concerning slaughter, transport, captivity, farming and zoos.
Mamak also sees more fundamental changes – new rights and legal categories for animals – as entirely “conceivable”, and even a future where courts or legislators may have to take into account AI-mediated animal experience testimonies. “The law has always depended partly on what we know about animals,” he says. “If technology radically changes what we can know about them, it may eventually change what the law believes we owe them.” Despite his positive outlook for the future, Mamak believes that just because AI can potentially help animals, it doesn’t mean it should be scaled without consideration of the environmental cost. Also, he cautions against the rise of a new kind of “ethics washing”, where companies pretend they are “talking to animals” about products while continuing to damage animal habitats.
We only need to look at the difference in how the Western world treats dogs and pigs to see how cuteness influences which animals we deem worthy of moral consideration. This is not a bias that AI can remove instantly. It is an inconsistency, however, that Mamak hopes animal communication would eventually make it harder to maintain. Rytis Maskeliūnas, a professor at the faculty of informatics at Kaunas University of Technology, and his colleagues analysed recordings of German shepherds before assigning four different meanings to their barks, relative to different dog “statuses”. It was a small data set, but he says it was a success. “We assume [dogs] don’t have a large vocabulary,” he says. “But if we can move beyond this, and incorporate the context, maybe we can assign more statuses of what they want to express.”
As more data – including monitoring apps and cameras – becomes available for AI to help track, scour through and attempt to decode the needs of animals, Maskeliūnas, like many others in the field, hopes it will one day help people to become better animal caretakers. The question becomes, however, that humans get the chance to understand animals at a deeper level, will we care what they have to say? Not just our pets, but the pigs waiting for slaughter in factory farms, or the whales swimming thousands of miles away in the ocean. Especially because it is highly unlikely that we’re going to like what the most mistreated animals in the world have to say.
But the question is: if humans get the chance to understand animals at a deeper level – and not just our pets, but the pigs waiting for slaughter in factory farms or the whales simming thousands of miles away in the ocean – will we like what they have to say?