A pretty big distance.
If a mobster threatens to burn down a building unless you buy their "insurance", that's a protection racket.
If someone finds a major fire code violation and threatens to tell the fire marshal about it unless they fix it within a certain timeframe, that's not a protection racket, even though there's technically a threat involved. If the building owner is a dick about it, then next time that person will probably just go directly to the fire marshal.
Plus, if the attack surface is huge and/or fractal, you will never run out of exploits. The more you pay people to find them, the harder they look...
That's not what happened here and isn't usually what happens, though? The reporter usually gives a timeline for fixing the bug before reporting externally, and often extends that deadline if it's clear the Company is working on it. This is separate from bug bounty payments.
> The more you pay people to find them, the harder they look...
Yeah... that's the point...
Your own staff and vendors are creating security vulnerabilities, and you wisely run a bounty program, to detect and alert you. And you only pay when they find a problem. It can be very economical hedge against both mistakes and systemic dysfunction.
Also, if the researchers were criminally-inclined, they could make more money selling vulnerabilities to someone, not alerting you.
> Nah, let's pay him instead!
is a solution, but obviously can't be the solution. From a distance, white hat "vulnerability disclosures" start to look like a protection racket.