What To Know
- An Australian artificial intelligence startup is turning the tables on cybercriminals by deploying thousands of undercover AI agents that deliberately engage with scammers, waste their time, and secretly collect valuable intelligence.
- A bank could instruct an agent to uncover the account number associated with a suspected investment scam.
An Australian artificial intelligence startup is turning the tables on cybercriminals by deploying thousands of undercover AI agents that deliberately engage with scammers, waste their time, and secretly collect valuable intelligence. Instead of waiting for unsuspecting consumers to lose money, Sydney-based Apate is sending convincing digital personalities directly into fraudulent operations, potentially transforming how banks, telecommunications providers, and governments combat financial crime.
Image Credit: Apate
Founded in 2024 following research conducted at Macquarie University in Sydney, Apate.ai has developed autonomous agents capable of answering suspicious telephone calls, responding to fraudulent messages, following phishing links, and infiltrating questionable online communities. As this AI Startups news report reveals, the company’s unconventional strategy combines deception, conversational intelligence, and large-scale automation to expose criminal networks before they successfully exploit legitimate customers.
Thousands of Digital Decoys Are Keeping Scammers Busy
Unlike conventional fraud detection systems that examine suspicious transactions or block questionable telephone numbers, Apate actively approaches criminals using AI-generated identities.
Its agents impersonate ordinary individuals with different ages, genders, accents, languages, and conversational styles. These fabricated personalities engage fraudsters in seemingly authentic conversations, encouraging them to continue revealing information about their operations.
The company reports that its agents have participated in approximately 2.5 million conversations, collectively consuming the equivalent of 17 years of scammers’ time.
However, frustrating criminals represents only one component of the strategy.
During these interactions, the agents attempt to identify bank accounts, cryptocurrency wallet addresses, telephone numbers, fraudulent websites, impersonated brands, and other operational details.
Such intelligence can help financial institutions and telecommunications companies recognize emerging threats before criminals successfully target customers.
A Family Picnic Inspired an Unusual Cybersecurity Business
The company’s origins can be traced to an unexpected encounter involving founder and CEO Dali Kaafar.
Several years ago, Kaafar received a suspicious telephone call while attending a family picnic. Recognizing the caller as a scammer, he decided to entertain his children by pretending to believe the fraudulent story.
The conversation continued for an extraordinary 44 minutes.
Although the incident amused his family, Kaafar recognized something important: every minute the scammer spent speaking with him represented time unavailable for targeting another potential victim.
He also realized that manually entertaining criminals was hardly an efficient use of his own time.
That experience inspired research at Macquarie University involving machine learning, natural language processing, and automated analysis of fraudulent conversations.
The research eventually evolved into Apate, named after the Greek goddess associated with deception.
Artificial Intelligence Is Becoming a Weapon Against Fraud
The startup’s emergence comes as criminals increasingly exploit artificial intelligence to conduct sophisticated financial scams.
According to the FBI’s 2025 Internet Crime Complaint Center report, authorities received more than 22,000 complaints involving AI-related crimes, with reported losses reaching approximately $893 million.
Investment fraud accounted for more than $632 million of those losses.
Criminals increasingly employ AI-generated conversations, synthetic voices, and deepfake impersonations of celebrities, business executives, and other trusted individuals to manipulate victims.
Apate’s technology attempts to counter these developments by applying similar advances in conversational AI against the criminals themselves.
The company’s agents combine advanced foundation models with proprietary fine-tuned systems trained using previous scam conversations and university research.
This approach allows the agents to maintain realistic interactions while identifying potentially useful intelligence within exchanges that might otherwise appear routine.
Banks Can Send AI Agents on Specific Intelligence Missions
Apate offers customers more than access to automated scam-baiting conversations.
Its intelligence platform provides analytics concerning fraudulent campaigns targeting particular businesses, customers, employees, products, and brands.
Organizations can also assign specific investigative objectives to individual agents.
For example, a bank could instruct an agent to uncover the account number associated with a suspected investment scam. Another agent might attempt to identify cryptocurrency wallets used by fraudulent trading platforms.
Text-based agents can investigate online forums, investment communities, and social media platforms before initiating conversations using fabricated identities.
Brad Joffe, Apate’s co-founder and chief commercial officer, explained to TechCrunch that traditional fraud prevention frequently intervenes after criminals have already approached potential victims.
Apate instead attempts to obtain actionable information directly from scammers, enabling customers to respond earlier.
This represents an important distinction between reactive fraud monitoring and proactive intelligence gathering.
Major Australian Companies Are Already Testing the Technology
Apate has attracted customers including Commonwealth Bank of Australia and telecommunications provider TPG Telecom, alongside government agencies, utilities, and other organizations.
According to Joffe, TPG Telecom has redirected nearly 590,000 suspicious calls to the company’s AI agents.
He also reported that Commonwealth Bank experienced a 76% reduction in publicly reported scam losses, although that improvement cannot automatically be attributed entirely to Apate’s technology.
Another unnamed Australian bank reportedly used the startup’s outbound agents to disrupt 22 separate scam campaigns.
These examples illustrate the potential commercial value of automated fraud intelligence, particularly when criminal networks operate across numerous communication channels simultaneously.
Nevertheless, independently establishing how much financial damage the technology prevents remains essential as adoption expands.
Investors Back Expansion into International Markets
Investor interest has accelerated alongside the company’s commercial development.
Apate secured $8.15 million in seed financing in August 2026 from investors including Lobby Capital, OIF Ventures, Investible, Concept Ventures, and Baobab Ventures.
The funding brought its total capital raised to just under $10 million. Operating with approximately 20 employees, the company intends to strengthen its technology while expanding internationally.
Its plans include establishing an office in London and increasing its presence in the United States.
Apate was also selected as a finalist in the Startup Battlefield 200 competition at TechCrunch Disrupt 2026 in San Francisco, providing an opportunity to showcase its approach before venture capital investors.
The exposure could support international partnerships as financial institutions search for more effective responses to increasingly automated criminal activity.
A New Front Line in the Global Fight Against Scams
For Thailand, where financial institutions, telecommunications operators, and regulators continue confronting digital fraud, Apate’s approach highlights an intriguing potential direction for cybersecurity investment.
Automated scam engagement could complement existing fraud prevention systems by identifying suspicious infrastructure and criminal payment channels before substantial losses occur.
However, deployment would require careful consideration of privacy protections, regulatory requirements, intelligence accuracy, and coordination between participating organizations.
The broader significance extends beyond keeping scammers occupied. By transforming fraudulent conversations into actionable intelligence, Apate is challenging the assumption that cybersecurity defenses must always wait for criminals to make the first damaging move.
Whether its model can deliver comparable results across different languages, jurisdictions, and criminal networks remains to be demonstrated. Nevertheless, the startup’s progress suggests that increasingly sophisticated AI agents could become valuable investigative partners in the international battle against financial fraud.
For more on Apate, visit:
