AI Trading Tools for Small Businesses: Separating Software From Snake Oil

Artificial intelligence has arrived in retail investing the way it arrived everywhere else, loudly, expensively, and with claims that outrun the evidence.

For business owners and professionals who manage their own investments between meetings, the appeal is obvious, a tireless assistant watching the markets while you watch the business. The trouble is telling the genuine tools from the merchandise.

Three Products Wearing One Label

Almost everything sold as AI trading is one of three things. Signal services scan markets and suggest trades, leaving execution to you, useful, but closer to a research subscription than automation.

Rule-based bots execute predefined strategies automatically, dependable in the way a spreadsheet is dependable, and about as intelligent. Genuinely adaptive systems, which adjust their approach as conditions change, exist but are rare at retail level, and rarer still with any published evidence they work.

Vendors know the third category sells, so the first two borrow its vocabulary. The subscription price tells you nothing, some of the dearest products are the least sophisticated.

The Evidence Standard

One question cuts through most of the marketing, can you show me live performance, not a backtest? Backtested charts are simulations against history, tuned until they fit, with costs modelled kindly.

Forward performance from a real funded account is the only number that counts, and the reluctance of most vendors to publish one is itself information. Fortunately, the verification is starting to be done independently.

UK-focused comparisons of the best AI trading apps now run these tools with real deposits and rank them on what actually happens, net returns after all costs, the honesty of the automation claims, and compatibility with FCA-regulated brokers.

That last point matters, because the regulatory protection in any automated setup lives at the broker, not in the software. Run whatever tools you like, but run them through an authorised venue.

Counting the Real Cost

The subscription is rarely the biggest expense. An automated system trading frequently pays the spread on every position, in both directions, plus overnight financing on anything leveraged and currency conversion on overseas markets.

A tool costing thirty pounds a month can quietly generate several times that in trading costs, all of it invisible on the vendor’s dashboard because it is charged by the broker underneath.

Before subscribing, sketch the maths, expected trades per month multiplied by typical spread, plus financing on the average holding period.

If the strategy needs to clear five percent a year just to pay its own running costs, the algorithm is working for everyone except you.

It is also worth asking who profits from the trading itself. Some free or cheap bots are distributed by brokers and introducing partners paid on volume, which means the software’s incentive is to trade often, not to trade well.

Frequent small trades look busy and feel productive, they are also exactly what a volume-paid distributor would want.

Genuine independence between the tool recommending trades and the venue charging for them is rarer than it should be, and worth confirming before you connect an account.

A Proportionate Approach

Treat AI trading tools the way you would treat any new supplier to your business. Trial with an amount you can afford to write off.

Total the full cost, subscription, spreads, any performance fee, and compare it honestly against a cheap tracker fund, which is the incumbent this technology has to beat.

Diarise a review at three and six months against that benchmark. And be unsentimental at the review, software that underperforms a passive alternative after costs is not an investment strategy, it is a subscription you forgot to cancel.

The technology will improve, and some of it is already useful for research and discipline. But in 2026 the trader’s edge is not owning the cleverest algorithm, it is applying ordinary commercial scepticism in a market that hopes you won’t.