One of the big things that artificial intelligence may be able to help us with, in general, is better analysis of massively complex systems that hide all kinds of fraud, smuggling, poaching and other malfeasance. Specifically, the technologies that we are discovering now will have big impacts on the biological world.

In something I wrote last week, I quoted one of our MIT people as suggesting that biology in the “operating system” of organisms. And that makes a lot of sense – recognizing that biology and human-made technologies are not mutually exclusive. But this intersectionality comes in all kinds of forms. One has to do with regulating wildlife.

What does this mean? Well, research shows that the United States, over the last 22 years, has imported some 2.85 billion individual animals of almost 30,000 species.

What do Americans do with all of this mail-order biodiversity? Some of these species become pets, or end up in zoos. Some contribute to research, or get made into consumer products, or medicines.

The idea, though, is that there’s so much of this trade that it becomes hard to monitor for safety, legality and sustainability.

Dangerous Species, Endangered Species

One problem with unregulated wildlife is that incoming species can pose some kind of danger to humans. In particular, officials are on the lookout for venomous species, although some admittedly fall through the cracks.

A resource from the Environmental Information Service shows officials trying to mitigate these arrivals.

“Prevention is the most effective and cost-efficient way of preventing those impacts that we know that nonnative species can have,” said Wesley Daniel, a U.S. Geological Survey research professional.

Among other illegal species are those species that are classified as endangered, having been poached, fished, hunted, or otherwise driven close to extinction.

That’s been a struggle for a long time: identifying and protecting these rare animals. And so much of this has to do with, again, monitoring complex systems.

In attempting to prevent the importation of problem species, officials are simply scrambling to find needles in haystacks of paperwork.

I heard a talk at Planet Action this year (disclaimer: I help to host these events) where Michael Tlusty, a Professor of Sustainability and Food Solutions at the University of Massachusetts Boston, talked about using AI to manage the seemingly endless lists of imports, to find those suspicious entries that might constitute policy violations.

One of the problems, Tlusty noted, is that importers have taken to using available broad codes like “tropical fish,” instead of actually listing the species. That understandably makes the monitor’s job a lot harder.

Showing off import sheets related to bats, seahorses and other creatures, Tlusty illustrated how this works. Of the seahorse, he said:

“Seahorses are CITES-listed species,” Tlusty noted, referring to the Convention on International Trade in Endangered Species of Wild Fauna and Flora . “They should have a permit. We should know where it came from. We should know how it was produced. There's a whole bunch of stuff we really need to know about it that we likely don't, because it's just hidden … in the paperwork.”

Understanding and Tracking Wildlife

Defining “wildlife” as “any non-domesticated animal,” Tlusty talked about how to improve these processes of evaluation.

“All of the seafood we bring in is wildlife,” he said. “And again, a lot of seafood is coming from developing countries, from small island nations. And this is where we need to make sure we're doing a good job and managing this well.”

Listing other kinds of methodology, Tlusty suggested that there’s still a lot of room for improvement with AI. The ENA approach, detection dogs and X-rays, he suggested, only go so far.

“All of these things work best if you're sending it to a target,” he said, going into a fundamental way to make X-ray analysis more efficient. “If you ask the X-ray, ‘what's in this box?’ It takes a lot of processing time and power to figure it out. If you ask the X-ray,’ is there a crocodile in this box? Is there a bird in that box?’ It's a much, much easier problem to solve.”

In the end, Tlusty said, it all comes down to being more deliberate.

“We need to understand the true nature of all the biodiversity that we're trading in,” he said in conclusion. “Data matter, and it matters to be able to find the seahorse. It matters to be able to find any one of the numerous species that are getting hidden in all of this trade that's happening. And it's important because biodiversity is important for communities. And where it's sustainable, we need to continue that. And where it's illegal, we need to find it and stop it.”

I think that improving these kinds of monitoring systems will have a positive impact on our world. And the needle in the haystack is less elusive when you have LLMs at your side. Stay tuned.