Bitget Blocks 150 Million Malicious Requests as $32.3M Recovered in Third Anti-Scam Push

Bitget is back at it. The exchange just launched year three of its Anti-Scam Month, dropping a full report on how fraud has mutated across crypto, tokenized assets, and other digital financial products. The report was put together with blockchain security firm SlowMist, and the numbers inside are pretty striking.

Between July 2025 and June 2026, Bitget’s security infrastructure intercepted over 150 million malicious requests. The team also flagged more than 13,000 high-risk IP addresses and worked through 18,135 user protection cases. Out of all that, $32.3 million tied to fraudulent activity was recovered. That’s not a rounding error — that’s real money that didn’t end up in scammers’ pockets.

Fraud Got Smarter, Fast

The stat that probably jumps out most: the share of active users trading across multiple asset classes went from under 1% in mid-2025 to over 10% by May 2026. That’s a massive behavioral shift, and fraudsters noticed. When users spread across more asset types, they open more surface area for manipulation. Scammers adapted accordingly.

The report, drawing on data from Bitget Research and SlowMist investigations, maps out how fraud campaigns are now built around complex narratives. It’s not just a fake website anymore. Scammers are combining AI-generated content, social engineering, and multiple communication channels — sometimes all at once — to build convincing traps. The goal is to make the lie hold together long enough to extract funds.

Some of the specific cases in the report are wild. One scam involved impersonating Cypriot President Nikos Christodoulides. Another — called the Truman Show scam — constructed nearly 90 fake investor identities to simulate an entire investment community around a target. That kind of operation takes coordination. It’s basically a production, not a phishing email.

And it doesn’t stop there. The report also calls out voice-cloning attacks, deepfake scams, AI-generated personas, malicious smart contracts, and synthetic investment communities. Each of these exploits a different vulnerability, and each one has gotten more believable as the underlying technology improves. Staying ahead of that is genuinely hard.

What Bitget Is Actually Doing About It

Bitget has been running Anti-Scam Month since 2024. Year one and two were about raising awareness. Year three is more operational — educational content, partnerships with industry stakeholders, and a continued collaboration with SlowMist that feeds directly into detection and recovery work.

More context: Polymarket Covers $3M in Losses After Phishing Attack Hits 4,500 Users

The exchange has a user base of over 125 million people. At that scale, security infrastructure can’t be reactive. You can’t wait for users to report a scam and then investigate. The 150 million intercepted malicious requests show what proactive looks like — catching threats before they reach users, not after.

The report also goes into victim psychology, which is probably the underrated part. Most security content focuses on the technical side: use two-factor authentication, check URLs, don’t click suspicious links. But Bitget’s report digs into how users actually get pulled in — the emotional dynamics, the common entry points, the moment where skepticism gets overridden. That’s harder to teach, but it’s probably where the real leverage is.

Practical guidance is included too. The report covers how to recognize AI-enabled scams, how to harden account security, and how to respond if something goes wrong. Clear, usable steps, not vague warnings.

The Bigger Picture

Crypto fraud isn’t a niche problem anymore. As more users move across multiple asset classes — and as the tools available to scammers get cheaper and more powerful — the attack surface keeps growing. AI-generated content used to require significant resources. Now it’s accessible. That changes the threat landscape in ways that are still playing out.

Bitget’s partnership with SlowMist gives the exchange access to on-chain investigation capabilities that most platforms can’t replicate internally. SlowMist has tracked some of the most complex fraud operations in the industry, and that expertise feeds directly into what Bitget can detect and recover.

See also: House Democrats Push SEC for Answers on AI Trading Advisers

The report also mentions Bitget’s partnership with UNICEF, targeting blockchain education for 1.1 million people. That’s a longer-term play — building baseline literacy around digital finance so that future users are harder to deceive from the start.

Fraud campaigns are getting more expensive to run, in theory, because they require more sophisticated tooling. But they’re also getting more effective when they work, because the assets involved are larger and the victims are more dispersed. The $32.3 million recovery figure is meaningful. But the 18,135 protection cases that generated it probably represent a fraction of what’s actually out there across the broader market.

Bitget’s user base spans over 125 million individuals globally, and the exchange says it intercepted those 150 million malicious requests while maintaining normal platform operations throughout the period.