The Problem With Crypto Tools That Promise Too Much

"Tell us how much money you want to make and let our bots figure out the rest for you." That line sat on a crypto trading company's public website, according to an SEC complaint, and the bots behind it didn't work the way investors were told. It's an extreme case, but the sales pitch is everywhere: crypto tools that promise too much, from bots with perfect win rates to "AI" that supposedly sees the next move first. Here's what regulators found when they looked behind those promises, why overpromising does damage even when the tool is real, and how I test a tool's claims before I trust it. Sources are linked so you can check me.

What does "promising too much" look like?

The CFTC, the US regulator for futures and commodities markets, put out a customer advisory whose title says most of it: AI won't turn bots into money machines. It describes scammers claiming their algorithms can generate returns of "tens of thousands of percent" or "100 percent 'win' rates," for bots that trade automatically and for signal services that send buy and sell calls. Its blunt line: "AI technology can't predict the future or sudden market changes."

In practice, the overpromise usually shows up as one of these:

• A guarantee. Fixed returns, or returns someone swears are certain, on something as volatile as crypto. Nobody can honestly offer that.
• A perfect record. 100% win rates, no losing weeks, screenshots of only green.
• Magic tech words. AI, neural nets, "proprietary algorithm," used as a reason to stop asking how it works.
• No downside. "Limited risk" or a hard loss cap that nobody can show you in practice.

What was actually behind the biggest promises?

The CFTC's own case study is Mirror Trading International. Per the advisory, Cornelius Johannes Steynberg took more than $1.7 billion in bitcoin from at least 23,000 people over about three years. The pitch was a "proprietary bot trading program" that came with a guarantee of at least 10% a month, with "no trading experience required" and buy-ins as small as $100 in bitcoin. In reality, very little money was traded. It ran as a Ponzi scheme, and he used MetaTrader demo accounts to create fake customer balances.

The line I opened with comes from a newer case. In an SEC complaint against Nathan Fuller and his company Privvy Investments, the SEC alleges he raised about $12.3 million from about 150 investors by claiming AI-based bots did high-frequency arbitrage across crypto platforms. The complaint says he promised profits of more than 100% in as little as 21 days, framed as a guarantee, and told investors the bots had stop-loss code capping losses at 3%. According to the SEC, the bots didn't work as described, and to the extent they worked at all, the code had no stop-loss or AI functionality. Those are the SEC's allegations, not a court's findings, but the pattern matches the CFTC's warning almost word for word.

Is it only outright scams?

No, and this is the part that matters for the whole industry. In March 2024, the SEC settled charges against two registered investment advisers, Delphia and Global Predictions, for claiming they used AI in ways they didn't. Delphia said its AI could "predict which companies and trends are about to make it big." Global Predictions called itself the "first regulated AI financial advisor." The firms paid $400,000 in combined penalties without admitting or denying the findings. The SEC's name for it is "AI washing."

Neither firm was a crypto bot, and nobody's saying they ran a Ponzi. That's exactly the point. Overpromising isn't only something fraudsters do. Regular companies do it too, because a bigger claim gets more clicks than an honest one.

Why overpromising hurts even when the tool is real

Most crypto tools aren't scams. Charting apps, explorers, wallet trackers and risk scanners do real work. But a real tool that oversells itself still causes damage, just more quietly.

• False confidence. A scanner that prints a big green "SAFE" teaches you to stop looking. A clean read only means no red flag was found in the data it checked, at that moment.
• Hidden costs. The CFTC tells people to factor in fees, spreads and subscription costs. A bot that wins small but pays fees on every trade can lose overall while showing you a high win rate.
• Bigger bets. If you believe a tool caps your downside, you size bigger. When the cap turns out to be marketing, the loss is bigger too.
• Erased trust. Every inflated claim makes the next honest builder harder to believe. That's a cost the whole space pays.

How to test a crypto tool's claims before you trust it

The CFTC's checklist is a good start: research the company and the people behind it (including a reverse image search on their photos), check how old the website's domain is at lookup.icann.org, get a second opinion, and know the costs. Here's how I'd add to it for crypto tools specifically:

• Look for the limits section. A serious tool tells you what it can't do. If you can't find a single caveat, that tells you something.
• Ask for proof, not a score. Can you click through to the transaction or record behind a flag? If it only gives you a verdict, you're trusting the marketing.
• Try it free with small stakes. Use a demo, a read-only mode or a tiny amount first. Watch how it handles a loss, not just a win.
• Check what it needs from you. A research tool shouldn't need your seed phrase. Anything that asks for it is a hard no.
• Translate the claim. A guarantee of returns, "no risk" and "AI predicts" all mean the same thing: walk away.

Where my own line is

I build crypto tools, so this cuts both ways, and I'd rather say that out loud. Honest marketing is a disadvantage in a feed full of 100% win rates. A tool that says "here's what I found, here's what I couldn't check" will always look less exciting than one that says "safe" in giant letters.

I'm taking that tradeoff on purpose. The free Interrogator links every flag to the record it came from and has a limits section that says data is a snapshot. That won't win the hype contest, but it's the only kind of tool I'd want pointed at my own money.

Fair warning about this article too: enforcement cases show the worst end of the market, not the average tool. Plenty of bots and scanners are built by people who are careful about what they claim. The goal isn't to distrust everything. It's to make tools earn trust with proof.

Where to go from here

For the bigger picture on who usually wins in the trenches, read why most Solana meme coin traders lose money. For signals you can check yourself without trusting anyone's score, see how to identify suspicious activity in new token launches and the guide on how to spot a rug pull.

The takeaway: the size of a tool's promise isn't evidence of anything except its marketing budget. A good crypto tool shows its work, admits its limits and leaves the decision with you. If it guarantees returns, it's telling you exactly what it is. What's the biggest promise you've seen a crypto tool make?

Still building. Still here. 🫡