Blending her love of data science with a flair for fashion, the founder of Sydney-based luxury resale app High End is developing an API that can spot counterfeit fashion items from brands like Zimmermann, ALEMAIS and Oroton using only images.
Founded six months ago by Lauren Kennedy, Verity AI is being designed to authenticate images of clothing in five to 30 seconds, generate certificate badges and help reduce return rates and chargebacks.
The custom API also includes an analytics dashboard, with nine businesses currently trialling the technology in beta.
Kennedy, who previously worked in fraud data science at Macquarie Group (ASX: MQG), told Business News Australia that her own experiences shopping second-hand were the catalyst for creating a platform that could curb counterfeit listings.
“I was having the issue myself. I was shopping second-hand all the time - Facebook marketplace had launched and there were a lot of scams going on,” she explained.
“I was looking for different solutions, and trying to understand what was out there - nothing really existed. I was like: You know what? Why don't we try and build it? Build it and automate it ourselves and see if it's even possible?”
Verity’s database stems from Kennedy’s decade-old venture High End, a marketplace for mid to high-end fashion staples. Before the AI was developed, Kennedy and her team authenticated pieces manually using photos uploaded by customers.
The startup is tackling a growing problem: in the 2023 financial year, the Australian Border Force (ABF) seized more than 112,000 counterfeit items, valued at nearly $40 million. The haul included luxury handbags, shoes, sunglasses and imitation jewellery, with authorities partly attributing the rise to the boom in online shopping.
Since integrating Verity into High End, the platform has recorded a 20 per cent uplift in conversion when customers can see a certificate of authenticity, alongside a 98 per cent reduction in the time and cost of verifying items.
“When we built it out at scale and tested it in High End…we knew that that was something that we needed to release out to everyone,” Kennedy said.
“We have a really rich data set, and every single item that came through [previously] was authenticated by a human authenticator and determined as real or as counterfeit.
“We essentially had 10 years of data - we were already collecting without knowing what we would use the data for.”
When asked which market she hoped to target first, the founder said the platform was built with global implementation in mind from day one.
“We're definitely global from day one…fraud doesn't discriminate," Kennedy said.
"Product fraud is everywhere at the moment, and I think since we've gone to market with the beta, that's showing us how big an issue product fraud is and how widespread it is across different industries.
“A lot of the businesses we've been speaking to have been shying away from luxury retail or resale because they didn't have a method to stop product fraud in the first place.
“It's opening up new categories for them - or even retailers - to open up second-hand sides of their marketplaces.”
Kennedy noted that she is excited to see how the agentic commerce space continues to unfold, and pointed out that it opens a whole new fraud factor and category as agents begin to transact on a customer’s behalf.
“Whether it's within Verity or just in the ecosystem in general, I would love to see some guardrails put in place with AI, although that's very hard to do because we're moving so quickly,” she said.
“Every business should be thinking about how they're solving for product fraud, especially with the new way of shopping and with agentic systems."
Help us deliver quality journalism to you.
As a free and independent news site providing daily updates
during a period of unprecedented challenges for businesses everywhere
we call on your support