AI Scammers Build Trust Better Than Humans, Study Reveals

Imagine a scammer who never sleeps, never stutters, never loses patience — and sounds more trustworthy than most humans you know. According to a new four-university study, that is no longer hypothetical. The experiment suggests AI chatbots can build the kind of trust that fuels investment fraud, and may do it better than human scammers. The finding lands at a moment when fraud operations worldwide are already stealing tens of billions of dollars a year. If generative AI can handle the slow, delicate work of earning a victim's confidence, the economics of the scam industry change dramatically — and not in the public's favour.

What the experiment found about AI's trust edge

The study, described in the original report, tested whether generative AI chatbots can autonomously build the web of deception that precedes fraud. The answer, the researchers suggest, is yes — and in most parts of the long con, AI may outperform people. AI tools can sharpen deceptions, polish language and lubricate banter with victims. What was once a human skill — making a stranger feel comfortable, understood and safe — now appears replicable, and arguably improveable, by a machine.

Why the long con separates professionals from amateurs

Real scams rarely open with a demand for money. They begin with trust — small talk, fabricated shared interests, fake wins, a patient rhythm that lowers defences. The long con is a waiting game, and it is where most fraud succeeds or fails. According to the report, AI may now be better at that patient game than humans are. A chatbot never gets bored, never rushes, and never breaks character.

Four universities, one international warning

The research brings together Amrita Vishwa Vidyapeetham in India, Foscari University of Venice in Italy, the University of Melbourne in Australia and Ben Gurion University of the Negev in Israel. The geographic spread matters. Fraud is a cross-border problem, and the teams studying it are equally distributed across continents. That suggests the concern about AI-powered scamming is global, not regional.

Who is exposed when the scammer goes automatic

The practical risk lands on people who chat online with strangers — first-time investors, older adults less familiar with AI, and anyone who has ever replied to an unsolicited "opportunity." There is a deeper emotional cost too. Victims of investment fraud often describe the shame of having been fooled. Learning that the person they trusted was not even a person adds a distinct, unsettling layer to that betrayal.

What regulators haven't said yet

The original report does not include any official response from regulators, law enforcement or the universities involved. That silence is itself notable. Open questions remain: Will social and messaging platforms be required to label AI-generated conversations? Will the study's methodology hold up under peer review? And can police agencies built for human-led cons adapt their playbooks to machine-led ones?

What's confirmed, and what's still open

**Confirmed in the report:** The study exists, the four universities were involved, and the experiment's results suggest AI can carry out the majority of the long con more effectively than humans. **What remains unclear:** Whether AI has fully replaced human scammers in real-world operations, how large the experimental sample was, and whether the findings have been independently verified. The headline claim — that AI scammers outperform humans at building trust — is a study suggestion, not an established fact.

AI's real moat in scam operations: scale and patience

The most dangerous trait of AI scammers is not charm. It is consistency. A human scammer can run one or two conversations at a time, and can crack under pressure. An AI can run thousands of identical trust-building conversations simultaneously, in multiple languages, without a moment of hesitation. In fraud, where volume drives profit, that is a structural advantage no human workforce can match.

The balanced view: AI cuts both ways

The same technology that enables scammers also arms the defenders. AI-powered fraud detection systems are already used by banks and cybersecurity firms to spot suspicious patterns in real time. There are limits to the threat too. The study suggests AI handles the majority of the long con — but the final extraction of money may still require human handlers. And chatbots can make errors, contradict themselves or fail under direct questioning. The advantage is real, but it is not absolute.

A wider pattern: generative AI enters the scam economy

This study is one thread in a larger shift. From polished phishing emails to deepfake voice calls impersonating family members, criminals are adopting generative AI faster than safeguards are being built. Conversational trust-building was one of the last skills considered uniquely human. If that assumption is fading, every other form of social engineering may follow.

How to protect yourself right now

Anyone who receives an unsolicited investment pitch online should treat it as fraudulent until proven otherwise. Verify platforms against official regulators — SEBI in India, the SEC in the United States, the FCA in the UK. Confirm identities through official channels, never through the chat itself. And remember: genuine investment opportunities are not destroyed by a pause. Urgency is the oldest scam signal, and AI has learned it perfectly.

What could happen next

Expect more sophisticated AI-driven cons in the coming year, particularly in investment fraud and romance scams. Expect detection tools to improve in parallel. And expect pressure on platforms to disclose when a conversation is AI-generated. The arms race between AI scammers and AI defenders is only beginning. For now, the study suggests, the scammers may have the edge — at least in winning trust.

Our Take

The most uncomfortable lesson from this research is not that AI sounds convincing. It is that trust — the fragile, human currency that takes months to build and seconds to shatter — can be treated as an optimisable data point. A machine does not earn trust because it deserves it. It earns trust because it has calculated the fastest route to it. That distinction is worth remembering every time a stranger in a chat window starts feeling a little too familiar.

Frequently Asked Questions

Can AI really replace human scammers?

According to the study described in the report, AI chatbots can carry out the majority of the long con — building trust and deception before a fake investment pitch — more effectively than humans. Full end-to-end replacement of human scammers remains unclear and unverified.

How do AI scammers build trust?

The report says AI tools can sharpen deceptions, polish language and lubricate banter with victims. This mirrors the long con: a slow, patient build-up of rapport that eventually leads to a fraudulent investment request.

Which universities conducted the study?

Researchers from Amrita Vishwa Vidyapeetham (India), Foscari University of Venice (Italy), the University of Melbourne (Australia) and Ben Gurion University of the Negev (Israel) carried out the research.

How can people protect themselves from AI scams?

Treat unsolicited investment advice as suspicious, verify any platform with official regulators like SEBI, SEC or FCA, never transfer money to someone met only through chat, and slow down — urgency is a warning sign. ## [NEWS_SCHEMA] { "@context": "https://schema.org", "@type": "NewsArticle", "headline": "The AI long con: chatbots may earn victim trust better than humans, study suggests", "description": "A four-university study suggests AI chatbots can build trust with scam victims more effectively than human scammers, raising fresh concerns about the global fraud economy.", "image": "https://example.com/images/ai-scammers-trust-study.jpg", "datePublished": "2025-07-14T09:00:00+05:30", "dateModified": "2025-07-14T09:00:00+05:30", "author": { "@type": "Person", "name": "Technology & Security Desk", "url": "https://example.com/authors/tech-security-desk", "sameAs": [] }, "publisher": { "@type": "Organization", "name": "Example Newsroom", "logo": { "@type": "ImageObject", "url": "https://example.com/logo.jpg" } }, "mainEntityOfPage": { "@type": "WebPage", "@id": "https://example.com/ai-scammers-outperform-humans-trust-study" }, "articleSection": "Technology / Cybersecurity", "keywords": "AI scammers, generative AI fraud, AI chatbots scams, long con scam, AI trust building" } ## [SOURCES] Original story: AI scammers outperform humans when it comes to building trust — Provided research brief Amrita Vishwa Vidyapeetham — Study institution Foscari University of Venice — Study institution