I'm afraid I don't have a link for this - the FT's web-site wouldn't let me find the story. One of their columnists suggested a way to beat spam: this is roughly the gist:

Each ISP identifies mail addressed to more than, say, 100 receipients. An employee retained for the purpose glances at the mail, and accepts or rejects the mail. This should be relatively easy, as most spam is readily spotted, compared to mailing list entries, etc.

The employee reviews, say, 1 a minute, for 7 hours a day, stopping 60 * 7 * 100 = 42,000 pieces of spam a day. Over a 200 day working year, and at a salary of say £20,000, that works out at about a quarter of a penny per spam stopped.

The author went on to suggest given the global volume of spam, only a couple of people would be needed to stop it all. This seems fallacious, as EACH ISP would need to employee their own blockers.

I think most people reading this site would see the implications for privacy - ISPs would (perhaps be legally required to) read any mail sent to large numbers of individuals. This is not something I'd look forward to.

Hopefully, the idea will die the death of a thousand rapidly knocked up columns, but it's worrying that privacy didn't even strike the writer (or his editor) as an issue. Particularly as, once the proposal had been brought in, there would be a natural pressure to reduce the number of receipients that "triggered" checking. Spammers would drop the size of their mailings, and so the checkers would have to look at an ever higher proportion of mails to have any effect.