Apate Deploys AI Agents Undercover to Fight Scammers

While many commercial enterprises deploy artificial intelligence to accelerate product delivery or streamline administrative workflows, one Australian venture takes a radically different approach. Apate.ai dispatches thousands of automated digital agents undercover with the explicit purpose of attracting fraudsters.

Established in 2024 following academic investigations at Macquarie University in Sydney, the enterprise utilizes a large fleet of virtual personas designed to mimic vulnerable individuals. According to TechCrunch reporting, these systems answer fraudulent phone calls, reply to suspicious emails, click through phishing URLs, and infiltrate online groups to engage perpetrators. The agents adopt diverse demographics, covering multiple age groups, genders, accents, and languages to prolong interactions with criminals.

TechCrunch notes that the company's bot network has collectively consumed 17 years of fraudsters' time across 2.5 million separate interactions. For institutional clients including financial institutions, telecommunications providers, public sector organizations, and utilities, the primary value lies in the actionable data extracted during these conversations. The undercover systems prompt fraudsters to disclose bank account details, cryptocurrency wallet addresses, telephone numbers, phishing domains, and impersonated corporate brands.

This intelligence is subsequently channeled into an analytics dashboard provided to clients, helping them refine their defensive strategies. Chief Commercial Officer Brad Joffe explained to TechCrunch that the startup combines frontier models with proprietary fine-tuned systems derived from founder Dali Kaafar's academic research. Kaafar originally conceived the concept after keeping an unknown scammer on the line for 44 minutes during a family gathering, which inspired him to automate the process.

Major industry players have already integrated the technology. Clients such as Commonwealth Bank of Australia, TPG Telecom, and various government bodies utilize the platform, with TPG routing nearly 590,000 calls to the startup's bots and Commonwealth Bank noting a significant reduction in public scam losses. Financially, the 20-person enterprise has secured nearly $10 million in total funding, closing an $8.15 million seed round in August with backing from Lobby Capital, OIF Ventures, Investible, Concept Ventures, and Baobab Ventures.

Why It Matters

The operational model demonstrated by Apate points toward a broader strategic shift in cybersecurity and fraud mitigation, moving organizations from reactive defense to proactive engagement. Traditional anti-fraud measures typically operate downstream, such as blocking phone numbers after an incident occurs or flagging suspicious transactions after funds have moved. By intercepting criminals at the earliest contact point through synthetic personas, organizations can preemptively harvest threat intelligence before end users suffer financial harm.

As generative artificial intelligence increasingly empowers malicious actors—with the FBI documenting tens of thousands of complaints and substantial financial losses tied to AI-driven scams—defenders must scale their countermeasures accordingly. Deploying autonomous conversational bots to waste perpetrator time and map out infrastructure changes the economic equation of online fraud, raising operational costs for criminal syndicates. However, as this adversarial loop intensifies, it also raises questions regarding how fraudsters might adapt their own verification methods to filter out automated entities, potentially leading to an ongoing technological escalation between synthetic defenders and digital threat actors.

Illustration: AI-generated image, not a photograph of the event.