{"id":100236,"date":"2026-10-11T06:21:54","date_gmt":"2026-10-11T10:21:54","guid":{"rendered":"https:\/\/overcentral.com\/en\/?p=100236"},"modified":"2026-10-11T06:21:54","modified_gmt":"2026-10-11T10:21:54","slug":"apate-ai-bots-trap-scammers-100236","status":"publish","type":"post","link":"https:\/\/overcentral.com\/en\/apate-ai-bots-trap-scammers-100236\/","title":{"rendered":"Apate\u2019s AI Bots Trap Scammers in Endless Phone Calls"},"content":{"rendered":"

For years, the narrative around artificial intelligence and cybercrime has been a grim one: AI supercharges phishing, generates hyper<\/a>-realistic deepfake voices, and writes convincing scam scripts at scale. But a new wave of countermeasures is turning that same technology into a weapon against the fraudsters themselves. Apate, an Australian company named after the Greek goddess of deception, has built a system of hundreds of thousands of AI-powered bots designed to do one thing: trap phone scammers in conversation for as long as possible, wasting their time while extracting valuable intelligence. It is a strategy of exhaustion rather than prosecution, and it is already being deployed by banks and telecom carriers around the world.<\/p>\n

Why Traditional Enforcement Cannot Stop the Scam Epidemic<\/h2>\n

The global crisis of online crime has long defied conventional law enforcement. Scammers operate across borders, often from jurisdictions where authorities have limited reach or little incentive to act. A call center running impersonation scams out of one country can target victims in a dozen others, and the proceeds can be laundered through a completely different financial system. Prosecutions are rare, resources are stretched, and the volume of fraudulent calls is staggering. This enforcement gap has created a vacuum, pushing governments, private companies, and security researchers to look for alternative solutions that do not rely on arresting their way out of the problem.<\/p>\n

One emerging approach is automation-based retaliation, or what might be called “scam-baiting at scale.” The idea is simple: if you cannot stop scammers from making calls, you can at least make those calls unproductive. By answering with AI, you deny scammers the human interaction they need to succeed, drain their time and operational capacity, and gather data that can be used to block future attacks. Apate is the most sophisticated example of this strategy in production today.<\/p>\n

How Apate’s Bot Army Wastes Scammers’ Time<\/h2>\n

Apate operates what its founder and CEO, Dali Kaafar, describes as a fleet of “perfect victims.” The system currently fields roughly 350,000 individual AI bots, each designed to mimic a real person who might be receptive to a scam call. When a scammer dials a number that has been flagged or fed into the system\u2014often numbers that have been recycled from known scam operations\u2014the bot picks up. It then engages the fraudster in a conversation that can stretch for minutes, sometimes longer, without ever actually falling for the pitch.<\/p>\n

The core insight is brutally practical. Scammers who use automated dialing tools can place thousands of calls per hour. Each bot conversation that lasts multiple minutes represents a significant chunk of that scammer’s available work time, time they cannot spend calling real potential victims. Kaafar puts it directly: “A minute that a scammer is talking to a bot or an agent is a minute where you’re probably saving hundreds, if not thousands, of possible people being reached out to by that exact same scammer.” The math is straightforward but devastating\u2014if a scammer has a fixed number of hours in a day, and a meaningful fraction of those hours are consumed by conversations that will never yield a payout, their entire operation becomes less viable.<\/p>\n

Intelligence Gathering as a Dual Purpose<\/h2>\n

Wasting scammers’ time is only half the mission. Every conversation Apate’s bots conduct is also an intelligence-gathering operation. The company reports that it has collected more than 250,000 discrete pieces of information about fraudsters in real time. These data points include scam URLs designed to harvest credentials, bank account details for money mule operations, phone numbers associated with fraud networks, and patterns of behavior that help identify new scams before they proliferate.<\/p>\n

This information is fed back to the banks and telecom companies that use the platform, allowing them to block fraudulent transactions, flag suspicious accounts, and warn potential targets before they are reached. The intelligence is also used to train the bots themselves, making them better at detecting new scam scripts and adapting to evolving tactics. In effect, Apate creates a feedback loop: the more time scammers spend talking to bots, the more the system learns about how they operate, and the better it becomes at frustrating them.<\/p>\n

Designing the “Perfect Victim”: Skepticism Without Suspicion<\/h2>\n

Building a bot that can hold a scammer’s attention without raising suspicion is a delicate engineering challenge. Scammers are trained to detect hesitation, overly robotic responses, or any sign that they are being baited. If a bot sounds too polished, too naive, or too obviously automated, the scammer will hang up and move on.<\/p>\n

Apate addresses this by giving each bot a distinct personality, language skill set, and background profile. Kaafar explains that the bots vary widely in their behavior. “Sometimes they do have WhatsApp, sometimes they don’t. Sometimes they pick up the phone, sometimes they just actually hang up on the scammer saying, ‘I’ll come back to you later,'” he says. The goal is to create a cast of characters that feel genuinely human, each with their own level of tech savviness, patience, and willingness to listen to a pitch.<\/p>\n

A critical design principle is that the bots must express a healthy amount of skepticism. If a victim agrees too quickly or too easily, a seasoned scammer will suspect a trap. Instead, Apate’s bots raise objections, ask clarifying questions, and push back in ways that feel natural. They leave enough openings, however, for the scammer to believe they can still close the deal. It is a tightrope walk between appearing gullible enough to pursue and smart enough to be believable.<\/p>\n

The result, as Kaafar describes it, is intensely frustrating for the scammer. The bot never quite takes the bait, but it never quite hangs up either. It keeps the scammer on the line, chasing a payout that will never materialize, while the system logs every scrap of data the fraudster reveals.<\/p>\n

What Happens When You Become the Scammer: A Firsthand Test<\/h2>\n

To understand how convincing these bots really are, the security newsletter Kernel Panic tested Apate using a demo version that flips the script: the user plays the role of the scammer, trying to persuade one of the AI “victims” to hand over money or personal information. The result was a surprisingly immersive and deeply irritating experience.<\/p>\n

Testers reported that the bot felt real enough to trigger genuine frustration. The AI persona expressed just enough curiosity and hesitation to make the user believe progress was possible, then subtly deflected every attempt to close. One moment the bot would seem to be considering the offer; the next it would ask a tangential question or express concern that derailed the pitch entirely. The conversation stretched on, and the scammer\u2014in this case, the person running the demo\u2014was left with the sinking feeling that they had been outsmarted by their own target.<\/p>\n

This is precisely the intended effect. Apate’s bots are designed to be “intensely frustrating” for scammers, a term that Kaafar uses deliberately. The goal is not just to waste time but to demoralize. A scammer who repeatedly invests ten or fifteen minutes in a call that yields nothing is less likely to maintain the persistence that fraud operations require.<\/p>\n

From Experiment to Enterprise: How Apate Scales<\/h2>\n

Apate is not a research project. It is a commercial platform backed by telecom infrastructure and deployed by financial institutions that face direct losses from authorized push payment fraud and impersonation scams. The company’s 350,000 bots represent a significant operational footprint, and the intelligence they generate is actionable in real time.<\/p>\n

Banks use the platform to monitor numbers that have been reported in connection with scams, routing incoming calls and messages to bots instead of allowing them to reach customers. Telecom operators, meanwhile, can deploy the bots at the network level, answering calls that originate from known high-risk sources or that match patterns associated with fraud. The system also extends beyond voice calls. Apate’s bots infiltrate scam chat groups on messaging platforms and respond to fraudulent text messages, applying the same logic across multiple communication channels.<\/p>\n

The company’s growth reflects a broader shift in how the security industry thinks about cybercrime. For years, the dominant paradigm was detection and remediation: find the scam after it happens, block the payment, and try to prosecute the perpetrators. Apate represents a preemptive approach, one that aims to change the economics of scamming by making it cost more time and yield less money.<\/p>\n

The Limitations and Risks of Automated Countermeasures<\/h2>\n

No single countermeasure is a silver bullet, and Apate’s approach has clear limitations. Sophisticated fraudsters who use targeted, manual attacks rather than mass automated dialing may be harder to trap. A scammer who calls only a handful of carefully selected victims per day may not be significantly impacted by a bot that wastes ten minutes of their time. Similarly, the system is most effective when it can anticipate which numbers scammers will call, which requires ongoing intelligence and cooperation from telecom providers.<\/p>\n

There is also the risk of adversarial adaptation. If Apate’s bots become widely known, scammers may develop methods to detect them, such as using speech analysis to identify unnatural pause patterns or asking questions that trip up the AI. Kaafar acknowledges this and says the system is designed to evolve continuously. The bots are updated based on the conversations they have, learning from each interaction to refine their responses and stay ahead of detection.<\/p>\n

Privacy considerations also arise. Apate’s bots need to sound like real people, and the conversations they conduct are recorded and analyzed. The company says it operates within legal frameworks and with the consent of its telecom partners, but the use of AI to impersonate potential victims raises questions that will only become more pressing as the technology scales.<\/p>\n

What Apate’s Success Means for the Future of Digital Security<\/h2>\n

Apate is part of a larger trend in which defenders are beginning to use the same tools that attackers have weaponized. Just as AI enables scammers to generate convincing scripts and deepfake voices, it also enables defenders to generate convincing decoys at a scale that was previously impossible. The asymmetry is shifting. A scammer who once faced only the occasional vigilant human target now faces an army of bots that can outlast them, outnumber them, and learn from every interaction.<\/p>\n

The implications extend beyond phone scams. The same principles\u2014automated deception, intelligence gathering through engagement, time-wasting as a defensive strategy\u2014can be applied to phishing emails, fake social media<\/a> accounts, and fraudulent online marketplaces. Apate’s model may well become a template for how organizations defend against a wide range of social engineering attacks in the coming years.<\/p>\n

What makes Apate’s approach particularly interesting is that it does not rely on legal or regulatory changes. It does not require international treaties or new law enforcement powers. It is a purely technical countermeasure that works within the existing infrastructure of the phone network. For banks and telecom companies that have struggled to protect customers from scams they cannot prosecute, that kind of practical, scalable solution has its own powerful appeal.<\/p>\n

The scammers will adapt, as they always do. But for now, an Australian company named after the goddess of deception is giving them a taste of their own medicine\u2014one endlessly frustrating phone call at a time.<\/p>\n","protected":false},"excerpt":{"rendered":"

For years, the narrative around artificial intelligence and cybercrime has been a grim one: AI supercharges phishing, generates hyper-realistic deepfake voices, and writes convincing scam scripts at scale. But a new wave of countermeasures is turning that same technology into a weapon against the fraudsters themselves. Apate, an Australian company named after the Greek goddess […]<\/p>\n","protected":false},"author":7,"featured_media":100238,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"fifu_image_url":"https:\/\/cards.overcentral.com\/cards\/en\/100236.png","fifu_image_alt":"Apate\u2019s AI Bots Trap Scammers in Endless Phone Calls","footnotes":""},"categories":[40668],"tags":[],"class_list":["post-100236","post","type-post","status-publish","format-standard","has-post-thumbnail","category-security"],"fifu_image_url":"https:\/\/cards.overcentral.com\/cards\/en\/100236.png","fifu_image_alt":"Apate\u2019s AI Bots Trap Scammers in Endless Phone Calls","_links":{"self":[{"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/posts\/100236","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/users\/7"}],"replies":[{"embeddable":true,"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/comments?post=100236"}],"version-history":[{"count":0,"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/posts\/100236\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/media\/100238"}],"wp:attachment":[{"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/media?parent=100236"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/categories?post=100236"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/overcentral.com\/en\/wp-json\/wp\/v2\/tags?post=100236"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}