Removing autoplay as the default setting on UK-licensed slot games cut the number of sessions running past 1am by 22%, according to operator-level data pooled from six mid-sized casinos over the twelve months to March 2025. The same dataset shows the average session length falling from 31 minutes to 24, with the sharpest drop among players aged 18 to 24. It is one of the cleaner natural experiments the sector has produced: a single interface change, applied at once, with no bonus or marketing attached.

The caveat matters before anyone gets carried away. This is not peer-reviewed, the operators involved won't be named, and the 22% figure describes sessions continuing past 1am, not harm prevented. But the direction of travel is consistent enough across the six books to be worth taking seriously, and it lands at an awkward moment for a Gambling Commission still weighing whether presets count as safer gambling tools or as marketing with a compliance badge.

What actually changed

Autoplay — the setting that lets a slot spin a set number of times without further input — was never banned. The Commission's remote customer interaction guidance, tightened through 2023 and 2024, pushed operators toward "friction" at the point of loss-chasing and extended play, and several large groups responded by flipping autoplay to off-by-default rather than removing it. Players can still switch it on. They just have to mean to.

That distinction is the whole point of the data. A hard ban would have produced a cleaner number but a worse comparison, because you couldn't separate the effect of losing the feature from the effect of a visible compliance intervention. Off-by-default keeps the feature available and changes only the effort required to reach it. In behavioural terms you're testing a nudge, not a prohibition, and the 22% is a nudge effect.

The numbers behind the headline

The pooled figures, as reported to us:

• Sessions active between 01:00 and 05:00: down 22% year on year
• Median session length: 31 minutes to 24
• Share of sessions exceeding 60 minutes: 18% to 11%
• Sessions where a deposit was made after 02:00: down 29%
• Autoplay re-enabled by players who found the setting: 41% of accounts, typically within the first three sessions

That last line is the one to sit with. Two in five players who went looking for autoplay turned it back on, and most did so almost immediately. The intervention isn't changing what those players do; it's changing what the other three in five do, and the 22% is largely their contribution. Whether that's a meaningful harm reduction or simply a redistribution of the same behaviour across a smaller group is the question the data can't answer.

There's also a selection problem worth flagging. Operators that flipped the default were, by and large, already investing in safer gambling tooling, and the comparison year includes a period of heavier bonus marketing across the market. Some of the overnight decline will be macro — a softer promotional environment, tighter affordability checks, the general drift of disposable income — rather than the toggle itself. A 22% figure with that much noise around it should be read as "somewhere in the high teens to mid-twenties," not as a precise measurement.

Why the overnight window is the one that matters

Session timing is a blunt proxy for harm, but it's a useful one. Overnight play correlates with a cluster of markers that operators' own models weight heavily: chasing losses after a bad evening, playing while intoxicated, playing alone when the household is asleep, and the long unbroken sessions that autoplay was designed to make frictionless. The Commission's own casework repeatedly features sessions running into the small hours as a common thread, and the 02:00 deposit decline of 29% is arguably a more interesting number than the session count, because it suggests the change is reaching people who were still putting money in.

It's worth being precise about what autoplay does mechanically. It doesn't increase stake or speed beyond what a player could achieve manually, but it removes the moment of decision. Every manual spin is a small opportunity to stop; autoplay converts a hundred of those into one. For most players that's convenience. For a player in a loss-chasing loop at 1:40am, it's the difference between a session that ends when they look at the balance and one that ends when the counter hits zero.

The 60-minute session figure — 18% down to 11% — is the cleanest expression of that. Long sessions didn't disappear, but the tail thinned considerably, and the tail is where the expensive outcomes live.

The counter-argument from the operator side

Not everyone in the industry reads the numbers this way. The argument you'll hear from product teams is that autoplay is a speed control, not a harm feature, and that removing it by default pushes players toward faster manual play or toward unlicensed sites that never changed anything. The second point is the harder one. The UK's black market is small relative to the licensed one but not negligible, and every friction point at a licensed operator is a marginal incentive to look elsewhere. There's no evidence in this dataset of a migration effect, but six operators over twelve months wouldn't detect one at the volumes involved.

The first point is more easily tested and largely answered: manual play did not speed up. Spin rates at the operators involved held roughly flat, and the share of players using turbo or quick-spin settings didn't rise meaningfully. Players who wanted speed found it; players who didn't were simply left alone.

What this means for the default-setting debate

The UK is moving, slowly, toward treating defaults as a regulatory instrument rather than a design choice. Deposit limits prompted at sign-up, loss limits suggested after a session marker is hit, and now autoplay off by default all sit in the same family: the operator chooses the starting position and the player opts out. The evidence here suggests that works, at least for the overnight tail, and that the effect is large enough to be worth the modest friction it imposes on everyone else.

The open question is whether the same logic survives contact with the things players actually value. Autoplay off is a cheap nudge because almost nobody chooses a slot on the strength of its autoplay setting. A default stake limit, or a default session timer that ends play rather than warning about it, would hit revenue in ways this doesn't, and the industry's willingness to accept friction tends to track how much friction costs. If the Commission extends the principle — and the consultation language suggests it might — the 22% will be cited approvingly by both sides, which usually means it's about to be asked to carry more weight than a single year of operator data can bear.

For players, the practical takeaway is smaller than the policy one. If you play slots overnight and you've noticed sessions running longer than you intended, the autoplay toggle is worth finding — not because a regulator says so, but because the moment of decision it restores is the one that tends to end the session. GamCare's helpline is on 0808 8020 133 if the sessions have stopped being something you choose.