Pet food and supplement companies that lack dog- and cat-specific research to support their claims risk losing consumer trust, as artificial intelligence tools make it easier for pet owners to check ingredient claims, according to Bill Bae, chief commercial officer of Radicle Science.
As AI-empowered pet owners demand proof over promises, Radicle Science is making gold-standard, pet-specific clinical trials faster, cheaper and more humane. Bae discussed the shift during an interview with Jackie Roembke for Petfood Forum Chats. Here are five takeaways for pet food professionals.
1. Consumers are using AI to fact-check pet product claims
Pet owners increasingly photograph products in stores or online and ask AI tools such as ChatGPT whether the products work, Bae said.
"What consumers are doing now — for themselves and for their pets — is taking a picture of a product in Costco or Petco or on an e-commerce platform, uploading it to ChatGPT and asking, 'Does this actually work?'" Bae said. "And for most products on the market today, the answer they're likely to get back is, 'No, there are no pet-specific studies.' Or if there is a pet clinical, it may have been conducted on 10 dogs rather than a statistically meaningful population."
2. Most pet products rely on "borrowed science"
Bae estimated that fewer than 10% of pet products have undergone a clinical trial specific to dogs or cats.
"I'd estimate less than 10% of pet products on the market have actually gone through a pet-specific clinical trial, in dogs or cats," Bae said. "Most are pointing to in vitro studies, rat studies or human studies and extrapolating from there. That's simply not the same thing."
3. Decentralized trials can cut cost and time
Radicle Science's Discovery platform conducts placebo-controlled clinical trials using pets living in normal home environments, rather than dogs housed in kennel colonies, Bae said.
Traditional labs or academic institutions can take eight months to a year just to begin a study, with no certainty on timing for results, he said. Radicle Science's approach — recruitment, biostatistics and data analysis handled in-house — turns results around in six to eight months, according to Bae.
On cost, Bae said traditional labs or academic institutions can run into seven figures, while Radicle Science's decentralized studies start at $70,000 for populations of up to hundreds of dogs.
4. Real-world study populations offer an advantage over lab conditions
Bae compared the approach to an automaker testing a vehicle on a controlled track versus one driven under everyday conditions in a major city.
"If Ford says a new Explorer completed 100,000 miles in a laboratory wind tunnel with no problems, versus Chevy saying a Tahoe completed 100,000 miles driven by a cab driver in New York City with no problems, most people are going with the Chevy," he said.
Bae said Radicle Science's panel of more than 100,000 Americans, over half of whom own pets, is built to reflect the broader U.S. population, allowing for more precise comparisons. "We can compare the GI system of a dog living in New York City to one living on a cattle farm in Texas, with a scalpel, not a broad blast," he said.
5. Formulating by extrapolation will not satisfy future scrutiny
Bae said many pet probiotic products still rely on strains studied for decades in other species.
"If you ask AI about a pet probiotic, and most of them are built around strains like E. faecium or Bacillus subtilis that have been around for 50 or 60 years, it will likely tell you the supporting research is outdated and not pet-specific," he said.
He said formulators often scale human dosing by body weight rather than generating pet-specific data. "A lot of formulating today still amounts to: if five grams works for a human, scale it by body weight for a 50-pound dog," Bae said. "There's math behind it. But math is not efficacy."
