Using artificial intelligence to fight wildlife crime more effectively

By Peter Kamocsai | January 2, 2020

In the two years since we examined a wildlife crime-fighting tool that uses AI, more countries have tested the AI tool, and researchers are helping to incorporate it into software used in thousands of national parks and protected areas around the world.

International wildlife crime is valued at $8 to $10 billion a year, according to the United Nations Office on Drugs and Crime. Yet catching or deterring thieves who harm animals and plant species is typically a cat-and-mouse game, and it is not easy for wildlife rangers to figure out where the next hit will happen. The AI-powered tool we wrote about in “The Future Has Begun: Using Artificial Intelligence to Transform Government” is proving to be one method for helping rangers outwit poachers.

The report, by the Partnership and the IBM Center for The Business of Government, featured the Protection Assistant for Wildlife Security, or PAWS, which uses data to predict where poachers will attack animals next and recommends randomized patrol routes to those areas.

In 2018, PAWS had been tested successfully in Uganda and in Malaysia. Since then, it has been tested in other countries, including Cambodia, where armed and organized poachers target animals such as wild deer and pigs. PAWS is also part of a larger puzzle in Cambodia: the World Wildlife Fund’s effort to reintroduce tigers to the region by 2022.

Tigers were last sighted in Cambodia in 2007 and conservationists believe they have gone extinct in this Southeast Asian nation. Their reintroduction is a complicated process, which includes making sure tigers have sustenance once they are back in the wild.

This is where PAWS comes in. If wildlife rangers use PAWS to protect the wild deer and pigs—the tigers’ prey—tigers might have a fighting chance of finding food on their own. And rangers could one day use PAWS to protect the tigers too.

Beyond Cambodia, researchers are working with wildlife organizations to integrate the AI-powered PAWS into non-AI software that rangers use. They hope to launch the improved software in up to 20 wildlife conversation areas in February 2020, which could expand to 600 areas by year’s end, according to the University of Southern California.

Protecting the environment does not stop with land animals. The PAWS team hopes to one day expand a version of PAWS to protect marine life and even forests.

But that is a story for another blog post, perhaps two years from now.

Interested in learning more about artificial intelligence and other new technologies in the federal government? Check out these blog posts:


Peter Kamocsai