How to Spot a Crypto Trading Bot Scam (And What Legitimate Actually Looks Like)

The dashboard showed 2.3% for the day. Below it, a running total, compounding since the account had been connected eleven months earlier. The bot doing the trading was called Royal Q, and for a long stretch, a lot of people believed that number was real money quietly growing in the background.

It wasn’t, not in the way the dashboard implied. Independent reviews of Royal Q found something else running underneath the trading claims: a six-tier referral structure, a 20% cut taken from every profitable trade, and a $120 activation fee before a single trade ever executed — the financial architecture of a recruitment scheme wearing a trading bot’s clothing. That’s the actual test for whether a crypto trading bot is legitimate. Not how confident the marketing sounds. What the numbers underneath it are actually built to reward.

Most articles on this topic hand you the same ten-item checklist — no guarantees, read the reviews, start small — and none of it is wrong, exactly. It’s just generic enough to describe almost any financial product, which makes it nearly useless as an actual filter. A sharper approach looks at a handful of well-documented cases, works out exactly what gave them away, and checks whether the pattern repeats. It usually does.

The Referral Math Is the Tell

Royal Q markets itself as an AI-driven trading bot, connecting to major exchanges through API keys and executing “quantitative strategies” on a user’s behalf. That part isn’t unusual — plenty of legitimate products describe themselves the same way. What sits underneath it is what separates a trading bot from a recruitment scheme dressed as one.

Independent reviews of the platform describe a six-level referral system, labeled V1 through V6, where earnings depend heavily on how many people a user recruits and how active that downline stays. Layered on top: a 20% commission on every profitable trade, a $120 activation fee just to start, and a separate withdrawal fee — a structure one reviewer flagged as significantly higher than what a standard exchange-hosted bot charges. None of that comes from trading performance. It comes from recruitment depth and transaction volume, which is a different business entirely.

A legitimate trading bot has one straightforward answer to “how do you make money”: a flat subscription, or a modest fee tied to your own account’s results. It has no structural reason to pay you for bringing in the next person, because your neighbor’s account has nothing to do with your bot’s trading logic. The moment a compensation plan depends on recruitment tiers instead of your own trading account, the word “bot” in the name is doing marketing work, not describing the mechanism.

There’s a second detail worth knowing about Royal Q specifically, because it shows up in a lot of these stories once someone looks closely: the company’s California LLC registration was marked terminated for inactivity in late 2022, and independent reviewers describe its regulatory licensing status as unclear at best. A trading product that’s been operating for years without a resolvable answer to “who regulates this, and where” isn’t a minor paperwork gap. It’s a structural absence of the accountability that makes any financial product trustworthy in the first place.

A bot that needs you to recruit isn’t confident enough in its own trading to let that be the entire pitch.

The Name on the Dashboard Is Often Borrowed, Not Owned

A specific kind of search brings people to this question — not “is automated trading safe,” but “is Royal Q legit,” or “is Gemini’s trading bot real.” That instinct, to check a specific name rather than trust a generic description, is exactly the right one. Scam operations know that a familiar name lowers a stranger’s guard faster than any feature list, so borrowing one is one of the oldest tricks available.

Royal Q itself claimed partnerships with major exchanges including Binance and Kraken — claims that, per independent review, were never independently verified anywhere outside the platform’s own marketing. That’s worth sitting with: a partnership claim that only exists on the page making it isn’t a partnership claim at all. It’s an unverified assertion borrowing a real company’s credibility.

The same logic applies to any “official trading bot” pitched under a well-known exchange’s name through a referral link or a Telegram group. Large, licensed exchanges generally don’t run their own guaranteed-return trading bot promoted this way — and if one ever did, it would be announced on the exchange’s own official channels, not discovered through a stranger’s referral code. Checking the exchange’s actual website directly, rather than trusting the name on someone else’s landing page, takes under a minute and closes this entire category of scam by itself.

A name you recognize isn’t a business relationship. It’s a word someone else chose to use.

What “Non-Custodial” Actually Protects You From

Underneath most of these schemes sits a second, separate question: who can actually move the money. This is a different axis from the referral-structure issue above — a platform can have a clean custody model and still be a recruitment scheme, or the reverse — which is exactly why it’s worth checking on its own.

A trade-only API key can open and close positions on an exchange account the user still fully controls. It cannot withdraw a single dollar, because the permission to do so was never granted. That’s a narrow, specific, checkable fact about a set of API permissions — not a promise about the operator’s intentions. The difference between a non-custodial bot and a custodial one comes down to exactly this: whether funds ever leave an exchange account the user controls, or move somewhere the operator controls instead.

If an onboarding flow asks for a wallet seed phrase, a deposit into an address the platform controls, or an API key with withdrawal permission switched on, the request itself is the red flag — no explanation afterward changes that. Setting up a trade-only key correctly takes about five minutes and permanently forecloses the worst-case outcome, regardless of which bot eventually connects to it.

Custody is the one variable that determines how a bad outcome actually plays out. A bad trading result costs money slowly. A custody failure costs everything, immediately, with no trade involved at all.

A Real Backtest Looks Boring, Not Impressive

Trading-bot marketing loves a screenshot: a dashboard full of green numbers, a chart climbing steadily up and to the right. It’s also the single easiest thing in the entire pitch to fabricate — a screenshot proves nothing except that someone can operate a screenshot tool.

A methodology, on the other hand, is checkable. Out-of-sample validation — splitting historical data into a research half and a completely sealed-off testing half, then checking a finished rule against that sealed-off portion exactly once — doesn’t produce a flashier chart. It produces a smaller, more honest one, because it eliminates the option to keep re-testing against the same data until the result looks good. That’s precisely why it’s rare in bot marketing: it’s a discipline that makes the numbers less impressive, not more, which is exactly the property a genuine methodology should have.

Ask what a bot’s backtest actually is — a picture, or a described, repeatable process — and the answer tells you more than the return figure ever will.

Where the Conversation Happens Is a Signal Too

One more pattern shows up often enough to be worth naming on its own: where a platform actually lets you reach it. Reviewers who tried to get real answers out of Royal Q describe email-only support with slow response times, and a community forum where a large share of posted questions simply never got a reply. That’s not an isolated customer-service complaint. It’s a structural choice.

Public, searchable support — a help center, a ticketing system with a track record, reviews that exist outside the platform’s own channels — creates a paper trail. A closed Telegram group or a WhatsApp chat run by an anonymous “account manager” doesn’t. Conversations there disappear, complaints don’t accumulate anywhere a new user can find them before signing up, and there’s no record for a regulator or a journalist to point to later. Funneling users into private, unindexed channels isn’t a support-style preference. It’s a way of making sure the previous person’s bad experience never reaches the next one.

A company confident in its own product wants its track record to be easy to find, not easy to bury.

The Guarantee Is Doing the Lying

Trend-following and other systematic strategies normally lose money on a large share of individual trades. That isn’t a malfunction; it’s the mechanism working as intended — most attempted moves fizzle out, a smaller number turn into real trends, and the strategy is built to stay in the game long enough to catch the ones that do. A strategy with no losing trades isn’t a better strategy. It’s a fabricated one.

So a guaranteed daily or monthly return is close to a self-disqualifying claim on its own. No systematic strategy, including honestly run ones, can promise a fixed positive number on a fixed schedule, because markets don’t cooperate with schedules. Real risk belongs in the same sentence as real potential, not buried in fine print several scrolls down — a genuine risk disclosure should read that way from the first paragraph, not just the last one.

A promise with no downside in it isn’t describing a trading strategy. It’s describing a sales pitch that hasn’t been asked a hard question yet.

This is worth separating clearly from the recruitment-and-custody problems above, because it’s a different failure mode entirely. A platform could have honest custody, no referral tiers, and a real support team — and still be dishonest about what the strategy itself can promise. Expectancy, not a guarantee, is the actual standard: a mix of many small losses and fewer larger wins, adding up to something positive over a long enough stretch, with no single day or week owed a specific number.

What Legitimate Actually Looks Like

Flip every pattern above and the honest version comes into view on its own. Fees tied to a user’s own account, not to how many other people that user brought in. A trade-only permission model instead of a request for deposits or seed phrases. A stated, specific methodology instead of a screenshot. A risk disclosure that shows up before anyone asks for it, not after.

None of that requires taking a marketing page’s word for anything — every one of those is a fact that can be checked directly, in minutes, before connecting an account to anything. The actual mechanism behind any systematic strategy worth trusting should hold up to exactly that kind of scrutiny, described at a level of honesty that doesn’t need a referral link to sound convincing.

The bots worth using are the ones with nothing to hide in how they get paid.