An Australian company just turned the tables on the global scam industry in spectacular fashion. Apate has deployed nearly 200,000 AI-generated personas that act as perfect, gullible marks. These bots answer phone calls, chat on WhatsApp and Telegram, and string fraudsters along for as long as possible while quietly harvesting intelligence that banks, telcos, and blockchain trackers can use to shut operations down.
One of the company’s monthly KPIs is how many F-bombs frustrated scammers drop on the “victims.” Founder Dali Kaafar isn’t kidding. “I think we’re the only company in the world that is actually keeping as part of their KPIs the number of F-words that scammers are dropping at them,” he said.
In a six-week stretch at the end of 2025, Apate’s bots handled 600,000 scam calls for just one Australian telecom, TPG. That burned more than 500 days of scammer time and was estimated to have saved around $13 million. The company now runs hundreds of thousands of calls and conversations daily across multiple channels.
From Family Picnic to Industrial-Scale Honeypot
Kaafar got the idea the hard way. In November 2021 he was on a picnic with his family in Sydney when a scam call came in. Instead of hanging up, he played along for 44 minutes. His kids loved the comedy show. He walked away thinking: if one person can waste a scammer’s afternoon for fun, technology can do it at industrial scale.
He was a professor at Macquarie University at the time. With help from doctoral students and early funding from Australia’s Office of National Intelligence, the project spun out into Apate in 2023. The company now works with most of the major Australian banks plus institutions in the UK, South Africa, and Southeast Asia. It also partners with blockchain analysis firms that want wallet addresses and money-movement intel before the cash disappears.
The bots started with 120 personas. They now number nearly 197,000, complete with accents, vocal tics, thinking noises, and distinct personalities. They were trained on hundreds of hours of real human scambaiter conversations so they know how to keep a con man talking without sounding like a script. On messaging platforms they act as honeypots, exploiting the same greed that drives the scams.
Not the Only Game in Town
Apate is the biggest player, but not the only one fighting fire with fire. British telecom O2 rolled out “Daisy,” an AI granny who loves talking about her cats, knitting, and family stories. She kept scammers on the line for up to 40–50 minutes at a stretch while the company turned the calls into public awareness content.
Big Tech is also swinging harder. Google sued a China-based operation it calls the “Outsider Enterprise” that allegedly used Gemini and other AI tools to power a phishing-as-a-service platform. The group was tied to more than 9,000 fake websites and over a million fraudulent URLs, including the familiar E-ZPass and government-looking text scams that hit American phones. Google is working with the FBI and the major U.S. carriers to dismantle the infrastructure.
OpenAI has been banning networks that use its models for scam scripts, translation, romance and investment pitches, and operational support, including operations linked to Cambodia and Nigeria. Meta has removed tens of millions of scam ads and accounts, and joined multi-agency disruption operations that took down large numbers of social media profiles tied to Southeast Asian scam compounds.
In March 2026, eleven major companies—including Google, Meta, Microsoft, Amazon, OpenAI, Adobe, LinkedIn, Pinterest, Match Group, Target, and Levi Strauss—signed the Industry Accord Against Online Scams and Fraud at a UN summit. The voluntary agreement commits them to share threat signals, coordinate defenses, and work more closely with law enforcement.
The Bigger Picture
Global scams are a trillion-dollar industry. Southeast Asian compounds, often involving human trafficking, industrialize the crime. American AI models and infrastructure have been abused to scale the operations—translation, persona generation, automated replies, and surveillance of workers. At the same time, defenders are using the same technology to fight back: conversational honeypots, real-time intelligence extraction, and rapid takedowns.
Apate’s researchers argue that defensive AI actually has a game-theory edge. The attacker’s bot is trying to get you to do something (send money, click a link, hand over credentials). The defender’s bot is trying to extract information. That asymmetry favors the good guys if the systems are built right.
This is not a complete solution as you all know. Scammers adapt. Deepfakes, voice cloning, and AI-generated personas are already in the wild. But industrial-scale time-wasting combined with intelligence gathering, platform enforcement, carrier blocking, and cross-industry information sharing is a serious upgrade from the old “report and pray” model.
Every hour a scammer spends arguing with an AI granny or a convincing middle-aged “victim” is an hour he is not calling someone’s actual grandmother. That is a win worth celebrating—and scaling.
—The Whatfinger News Team: Ben and Beth
Resources
- 200,000 Fake AI ‘Victims’ Deployed To Scam-Bait Online Fraudsters
- Fake AI ‘Victims’ Are Scam-Baiting 600,000 Fraudsters Every Month
- Apate | AI powered fraud prevention and intelligence
- Google sues suspected AI scam network
- How Google is combatting AI scams and dismantling the “Outsider Enterprise”
- Scam operations: Online fraud networks
- Fighting Scammers and Protecting People With New Technology and Partnerships
- Industry Accord Against Online Scams and Fraud
- O2 unveils Daisy, the AI granny wasting scammers’ time
- ‘Dear, did you say pastry?’: meet the ‘AI granny’ driving scammers up the wall
Top Features and Vids across all Whatfinger sites right now
-
- TREASON IN PLAIN SIGHT: Democrats Kept CCP Honeytrap Asset Eric Swalwell on the House Intelligence Committee After the FBI Warned Them
- Americans Are Finally Done Subsidizing the World’s Pharmacy Bills — Trump’s Most Favored Nation Plan Is Delivering the Biggest Prescription Drug Price Drop in Over 60 Years
- China’s Cash Pipeline to America’s Marxist Candidates: Why Are We Letting This Happen?
- They’re All for Socialism… Until It’s Their Own Grades
- Comcast turned millions of routers into motion sensors. The same WiFi signals that detect you moving through your living room can identify you with 99.5% accuracy.
Get on over to Whatfinger.com – Whatfinger News does millions of pageviews a month for a reason…. Truth, Justice, the American way, plus news on the Left as well – all links from all sites gives you the true news. – CLICK HERE 🛑
Drudge Report Alternative
This one is dark…. After the Democrats steal both Houses in November…
