The acceleration to digital as a result of the pandemic increased online shopping and auction fraud in the UK alone by 65%. With fraudsters using increasingly sophisticated means to scam users, we recognised that a new and more holistic approach was needed to tackle online threats effectively.
Pasabi’s behavioural analytics approach allows us to do a more extensive and comprehensive analysis of fraudulent activity online. Our powerful and highly scalable AI analytics platform tackles online infringements at scale. From our initial focus on counterfeit products and fake reviews detection, we have expanded our coverage for online platforms and brands to tackle online infringements with additional tools for unauthorised seller monitoring and content moderation.
How we do it
Using Natural Language Processing (NLP), machine learning classifiers, time series analysis, clustering technology and bespoke algorithms, Pasabi’s platform correlates thousands of data points. Our focus is the negative behaviours of bad actors, whether it’s counterfeit products, fake reviews, illegal content or unauthorised selling. Our approach recognises the ineffectiveness of the ‘whack-a-mole’ approach and instead looks for connections in the data to identify the worst offenders, surface networks working together across multiple channels and disrupting them at source. Pasabi’s analysis uncovers the scale of the problem and supports enforcement teams with actionable evidence for online removal and offline legal enforcement.
Chris Downie, CEO at Pasabi, explains:
“The move to develop additional tools to tackle online fraud felt like a natural and complementary extension of our existing offering. Pasabi’s mission has always been to use technology to make the internet a safer place. Our interest in this area led us to focus on the negative behaviours of bad actors across multiple channels and sectors to spot patterns in the data. This approach got us thinking about the other challenges brands and online platforms face and where our technology could help.
“With unauthorised selling, for example, the behaviours such users exhibit follow similar patterns. Our cluster detection technology highlights unusual behaviour and identifies groups working together, abusing brand agreements. With fraudsters continually trying to evade detection, brand owners and platforms can feel they’re fighting a losing battle. Pasabi’s technology captures visual evidence of products being added to baskets to track when and where products were available for sale to the public. So, even if an unauthorised seller takes their site down to ‘hide’ their actions, the evidence has been captured to allow enforcement teams to take the necessary measures to pursue enforcement action.
“As regards content moderation, casting a wider lens across the data and looking beneath the surface of what’s written allows us to uncover deeper insights to understand the actor’s intent. Our holistic approach enables us to more accurately detect repeat offenders instead of mistakenly identifying someone who’s speaking up against another user’s racist comments.”
Whether you’re a brand looking to scale and maximise your brand protection activities, or an online platform trying to safeguard your users from fraudulent activity, Pasabi’s behavioural analytics platform provides an effective, scalable and easily integrated solution, giving you and your customers the confidence to trust.