Apate deploys AI agents to bait scammers and gather fraud intelligence

Apate.ai, an Australian startup founded in 2024, is deploying thousands of artificial intelligence agents to actively engage with fraudsters. Rather than using AI solely for internal efficiency, the company sends these digital decoys undercover to answer scam calls, respond to phishing attempts, and join malicious online groups. The initiative aims to disrupt criminal operations by wasting scammers' time and gathering intelligence before human victims are harmed.

What happened

The concept originated from a personal experience involving Dali Kaafar, Apate’s founder and CEO. During a family picnic, Kaafar received a scam call and decided to play along, keeping the fraudster on the line for 44 minutes. He realized that this duration represented time the scammer could not spend targeting actual victims. This insight led to research at Macquarie University in Sydney, where Kaafar and his colleagues explored using machine learning and natural language processing to analyze scam content and improve detection methods. The resulting technology evolved into Apate, which now operates as a commercial entity.

Apate’s agents masquerade as potential victims with diverse demographics, including various ages, genders, accents, and languages. They interact with scammers across multiple channels, including voice calls, emails, text messages, and social media platforms. The primary goal is twofold: to consume the scammers' resources and to extract specific details about their operations. According to the company, its fleet has engaged in 2.5 million conversations, effectively wasting 17 years of cumulative scammer time. More critically, the agents solicit information such as bank account numbers, crypto wallet addresses, phone numbers, and malicious URLs, which are typically only discovered after a successful fraud.

The startup recently secured $8.15 million in seed funding in August from investors including Lobby Capital, OIF Ventures, Investible, Concept Ventures, and Baobab Ventures. With total funding just under $10 million, the 20-person team plans to expand internationally, with offices planned for London and increased presence in the United States. Apate is also a finalist in the Startup Battlefield 200 at TechCrunch Disrupt, where it is pitching to venture capital judges in San Francisco.

How it works

Apate combines frontier large language models with its own fine-tuned models to power its agent fleet. The system is trained on data derived from the conversations its agents have with scammers, as well as the foundational research conducted at Macquarie University. This hybrid approach allows the agents to maintain realistic personas and adapt to different scamming tactics dynamically. The agents are designed to be persistent, keeping fraudsters engaged long enough to reveal operational details.

Customers, which include banks, telecommunications companies, government agencies, and utilities, access an insights portal that provides analytics on scams targeting their specific user bases or brands. Organizations can assign specific missions to the agents, such as infiltrating investment groups on social media to uncover crypto wallet addresses used in fraud schemes. The text-based agents actively search forums and social platforms for suspicious activity, then enter conversations using fabricated identities to surface further intelligence. This proactive model contrasts with traditional reactive measures like blocking phone numbers or monitoring transactions after funds have been stolen.

Key details

• Apate was founded in 2024 based on research from Macquarie University in Sydney.
• The platform has conducted 2.5 million conversations, wasting an estimated 17 years of scammer time.
• Clients include Commonwealth Bank of Australia and TPG Telecom, with Commonwealth reporting a 76% drop in publicly reported scam losses.
• TPG Telecom has diverted nearly 590,000 calls to Apate’s agents to handle potential fraud interactions.
• The FBI recorded over 22,000 complaints of AI-related crimes in 2025, resulting in approximately $893 million in losses.
• Apate raised $8.15 million in seed funding in August, bringing its total raised to just under $10 million.

Why it matters

For software engineers and security leaders, Apate represents a shift from defensive postures to offensive intelligence gathering. Traditional anti-fraud systems often react after a breach or loss has occurred, such as blocking a phone number only after it has been used for thousands of scams. By engaging scammers directly, organizations can obtain actionable data—such as malicious domains or wallet addresses—before these tools are deployed against real customers. This proactive approach allows for faster updates to blacklists and fraud detection rules, potentially preventing losses rather than just mitigating them.

The rise of AI-enabled fraud makes this approach increasingly necessary. In 2025, AI-related crimes resulted in significant financial losses, with investment scams alone accounting for more than $632 million of the $893 million total reported to the FBI. Fraudsters are using AI to automate conversations and create deepfakes of trusted figures, making detection harder for humans. Apate’s use of AI to counter AI-driven fraud highlights an emerging arms race where automated agents must be deployed to match the scale and speed of modern cybercriminal operations.

What you can do

• Evaluate whether your current fraud detection systems rely solely on reactive measures like transaction monitoring or number blocking.
• Investigate platforms that offer proactive threat intelligence by engaging with bad actors to gather early indicators of compromise.
• Consider integrating external fraud intelligence feeds into your security operations center to update blocklists in near real-time.
• Review your organization’s exposure to AI-driven social engineering attacks, particularly in customer support and investment channels.
• Monitor emerging startups in the cybersecurity space that specialize in adversarial AI or automated deception technologies.
• Educate teams on the evolving tactics of scammers who use AI, ensuring internal awareness matches the sophistication of external threats.