The thing is that many of these people think they know what they are doing and do not think about security, only how awesome LLMs and AI make their experience until something bad happens.

Even though this was a valid critical bug [1], you need to enable screen sharing and allow connections to and from port 5900 on your router for a remote person to be able to exploit this.

Any security conscious person would probably be using a VPN (Wireguard or Tailscale) to prevent something like this, in the first place.

> almost every time it’s to click “OK” on a stupid prompt that I’ve long since stopped taking seriously.

The line above tells you how seriously this person takes security prompts.

[1]: https://nvd.nist.gov/vuln/detail/cve-2026-65400