A "dead man's switch" is a momentary switch designed to disengage if the person operating it dies or otherwise becomes incapacitated. The quintessential example is trains: the person operating the throttle must also engage a dead man's switch at all times, because if they were to have a sudden medical emergency, you don't want the train barrelling ahead with nobody at the helm. As soon as they release the switch, the fail-safe circuit is broken and the system automatically engages the emergency brake.
Computer systems can require users to keep a metaphorical "dead man's switch" active as a security precaution. One of the many government services I need to access for work uses this approach to prune inactive accounts: once a month, they'll send me an e-mail with a link I need to click to keep the account active. If I don't click it within 10 days, the account is proactively disabled.
This is obviously how all paid subscriptions should work by default, and this should be mandatory. There's no reason for subscriptions not to work like this other than the company's desire to exploit human fallibility. I think we've all had the experience at least once of signing up for Netflix or whatever, realizing that we haven't watched it in months, and sheepishly cancelling after spending dozens or even hundreds of dollars on absolutely nothing.
The default when you sign up for Netflix or anything should be that they send you an e-mail 3 days before the end of each service month with a link to keep your subscription active. If you don't click that link within 72 hours, your subscription would automatically be cancelled and you won't be charged for another month.
Subscription-based companies are very opposed to this idea, naturally, because much more of their revenue than they want to admit comes from inactive accounts that, nonetheless and for whatever reason, continue to be billed. Netflix would be much less dominant if the only people giving them money were the people who genuinely, actively want the product they're selling. But here's the thing, there's no reason it shouldn't work this way.
The companies would inevitably cite customer convenience as a reason this is a bad idea. I think the inconvenience of clicking a link every month is overblown, but even if you accept this argument, there's a very easy fix: let customers opt into autobilling. If someone really thinks clicking a link once a month is a huge burden, they can check a box in their account settings and not have to worry about it. Maybe they have so much money that the 60 seconds a month they would need to think about it is worth more than $20 to them. That's their prerogative.
You could also offer 3, 6 and 12-month billing options for customers who want the convenience but don't want to commit long-term. At the end of their billing cycle, send them two links: "Click here to renew your X-month subscription" and "Click here to renew for another month".
If you miss the window to renew, if the e-mail goes to your spam folder or you forget about it, then the horrifying outcome is... the next time you go to use the service, there's a banner saying "Your service is on hold, click here to subscribe for another month". You click the link. They have your credit card on file. They bill you for another month. You can immediately use the service again. You would barely even notice it. It would be no more friction than dismissing a cookie banner.
There's been a lot of talk about regulating and punishing companies that use "dark patterns" in subscription billing, such as plans that allow you to sign up for an account online but require you to make a phone call if you want to cancel. Or onerous multi-step cancellation procedures with unclear wording that tricks or frustrates people into staying subscribed. This kind of regulation is good, but I don't think it goes far enough. The monthly subscription as we understand it is, itself, a dark pattern.