Community Trust ScoreVerified
Scammers are losing. Not to cops, not to regulators — to bots that pretend to be gullible grandmothers and lonely retirees, burning through fraudsters’ time while quietly harvesting their secrets.
Australian tech firm Apate has built a fleet of 197,000 AI personas — each with a distinct voice, a distinct backstory, a distinct way of stumbling over words — and deployed them against the global scam industry. The bots run on phones, on WhatsApp, on Telegram. They chat. They hesitate. They say “oh really?” at exactly the right moment. And every hour a scammer spends sweet-talking one of these fake victims is an hour that scammer isn’t stealing from a real person. Apate tracks how often scammers lose their temper and resort to profanity. That’s basically their version of a KPI.
Six Weeks, 600,000 Calls, $13 Million Saved
The numbers get pretty wild fast. In the six weeks leading up to the end of 2025, Apate’s bots ran 600,000 scam calls on behalf of TPG, an Australian telecom provider. The company puts the estimated savings at $13 million — real money that didn’t end up in scammers’ pockets because those scammers were busy arguing with a machine that wouldn’t hang up.
Founder Dali Kaafar didn’t start with 197,000 bots. He started with 120 AI personas back in 2021, after a personal run-in with a scammer pushed him to do something about it. The company has expanded steadily since, training its models on extensive recorded conversations — real exchanges between scam baiters and fraudsters — to make the bots sound genuinely human. Pauses, filler words, the slight confusion of someone who kind of believes whatever they’re being told. It’s all there.
Apate now works with clients in Australia, Asia, Africa, and Europe. Banks and telecom companies are the main customers. They feed scam calls into the system and let the bots do the talking.
Crypto Wallets, Indian Bank Accounts, and Intelligence That Actually Matters
The phone-wasting part is almost the least interesting thing Apate does. What’s more valuable is the intelligence the bots pull out of scammers mid-conversation. Crypto wallet addresses. Payment routing details. Network contacts. Apate collaborates with a leading blockchain analysis firm — unnamed in the company’s materials — to track scammer wallet activity in real time, giving financial institutions a shot at blocking fraudulent transactions before money moves.
And it’s not just wallets. Apate’s bots have uncovered a marketplace offering verified Indian bank accounts, with commissions paid to brokers for scams — those commissions settled in USDT. That’s a pretty specific window into how fraud networks actually operate, the kind of thing a traditional investigation might take months to surface.
Scammers themselves are starting to use AI now. That’s the uncomfortable part. Apate knows it. Kaafar’s position, backed by the company’s own game-theory research, is that defensive bots still hold a structural edge — because the whole point of the interaction, from Apate’s side, is intelligence gathering. The scammer has to reveal something to run the con. The bot just has to keep them talking. That asymmetry probably won’t disappear even as scammer AI gets more sophisticated. Probably.
The company’s funding includes support from the Office of National Intelligence, which gives a sense of how seriously governments are starting to take this kind of counter-fraud infrastructure. It’s not a startup side project anymore.
A $1.24 Trillion Industry That Isn’t Slowing Down
The scale of what Apate is fighting is genuinely staggering. The global scam industry is worth an estimated $1.24 trillion — bigger than most national economies, and growing. Phone scams, crypto romance fraud, fake investment platforms, pig butchering schemes. The tactics keep shifting, and they’re getting harder to spot. Scammers now use AI-generated voices, deepfake video calls, and sophisticated social engineering scripts that can fool even cautious people.
Apate’s bots extend across messaging platforms, not just phone calls. WhatsApp and Telegram are active theaters. The bots act as decoys, pulling scammers deeper into fake conversations while logging everything. Banks use that data to flag accounts. Telecoms use it to block numbers. It’s a feedback loop that gets tighter the more calls the system handles.
Kaafar’s bet is that volume and intelligence beat sophistication. Keep the bots running, keep the data flowing, keep refining the models. The scammers swear at the bots more and more often. That’s probably a good sign.
Apate’s bots engaged in hundreds of thousands of interactions daily across multiple continents, with 197,000 active personas as of the company’s last published figure.
Frequently Asked Questions
How many AI personas does Apate currently operate?
Apate runs 197,000 AI personas, each with unique vocal characteristics, trained on real recorded conversations between scam baiters and fraudsters.
What did Apate’s bots uncover about scammer payment networks?
Apate’s AI found a marketplace offering verified Indian bank accounts to brokers, with commissions for scams paid out in USDT.
Why It Matters
The deployment of AI bots to combat scams highlights a significant shift in the tactics used to address fraud in the digital age. As the cryptocurrency and tech markets continue to grapple with the pervasive threat of scams, innovative solutions like those from Apate could reshape the landscape of consumer protection and trust. This initiative may also influence regulatory approaches and encourage further investment in AI-driven security measures across various sectors, emphasizing the importance of proactive defenses in an increasingly complex digital environment.
