This story in the New York Times tells of a year-old startup company called Social Intelligence, whose remit consists of assembling a dossier of information on job applicants based on their online activity over the past seven years.

Now, you may argue that employers are perfectly within their rights to carry out background checks on prospective employees, and you'd be absolutely right — it's why schools and other positions which place people in positions where they will be dealing with "vulnerable" individuals require a disclosure check to make sure the applicant doesn't have a checkered criminal past. Evidence of professional honours and charitable work also helps make an employer feel that not everything listed on a CV is a fabrication.

The concerning part is what else Social Intelligence looks for — according to the NYT article, "online evidence of racist remarks; references to drugs; sexually explicit photos, text messages or videos; flagrant displays of weapons or bombs and clearly identifiable violent activity." The concerning part is not the type of content that the company is looking for — it's how it might be stumbled across in a typical Internet search. That is, completely lacking in context. I'm not for a moment condoning violent activity, racism or anything else dodgy. But, frankly, everyone makes jokes, and sometimes those jokes are off-colour. Everyone has embarrassing Facebook photos, many of which are not what they seem. And if someone's had a puff of weed of a weekend and had a good giggle about it with their friends, that doesn't make them an inherently bad person, either.

"We are not detectives," said Max Drucker, CEO of the company. "All we assemble is what is publicly available on the Internet today."

Fair enough; but where does it stop? Once employers get the message that it's okay for companies like Social Intelligence to start trawling through your online background, what's to stop them from rejecting you based simply on something you said to your friends, or who you associate with online. This is particularly relevant given the "amusing" practice of friends "facejacking" or "fraping" each other's accounts given the opportunity — perhaps they left their account logged in, perhaps they left their phone on the table to go for a piss. Regardless of how or why it happened, a good-natured facejacking with all its usual excesses could well lead to someone's job prospects being dashed on the rocks — through no fault of the candidate.

Then there's the privacy question. Not necessarily the "what you share" question — that's a different matter entirely, and one which every individual must decide upon: what are you willing to tell people online? No, the privacy question I'm concerned about here is the divide between the personal and the professional. We're all different people at work — we behave in one way when we're on the clock, expected to be that person listed in the Person Specification and deal with customers and clients in the way we're supposed to, but as soon as 5pm rolls around we're off down the pub, swearing like a sailor, giving each other light-hearted ribbings and possibly making fools of ourselves. This latter part of the day doesn't affect our capability to do the job effectively. This latter part of the day is completely irrelevant to an employer — and, given most social networks' focus on the "personal" rather than the "professional", most social networks save the interminably boring LinkedIn are also completely irrelevant to an employer.

As someone who suffered workplace bullying from management partly as a result of some extremely vague negative comments on Twitter (which didn't mention the company in question at all, I hasten to add) — and witnessed several colleagues get fired over a Facebook prank that went awry — I feel particularly strongly about this. The things I said online were vague, not directed at my employer but at my life situation in general, and designed to let my friends who cared about me know how I was feeling — which wasn't great at the time. My professional life had no place intruding on my personal life — my personal life was not affecting my job performance, which had never been better. There were facts and figures and customer satisfaction surveys to prove it. Ironically, all the poor treatment I received at the hands of this shockingly bad management did was make me more likely to badmouth them now that I've left the company. But specifics of that are for another day.

The best analogy I can think of for Social Intelligence's work would be if as part of your job interview you had someone from the company follow you to the pub in the evening, follow you home, watch you go about your daily business, watch you have a shit, shower, shave, and then go through your bins just for good measure. In the days before social networking sites employers didn't do this, so just because there is the possibility for unprecedented invasions of privacy doesn't mean that it should happen.

Sadly, however, in the modern world, a lot of people seem to think that the words can and should are, in fact, interchangeable. And as such we end up with companies such as Social Intelligence rifling through candidates' virtual dirty laundry in an attempt to come up with the one tragic flaw that means Mr Perfect is not, in fact, quite so perfect for this position after all.

To me, the concept of "watch what you say" goes against everything social media — which should, in essence, be the ultimate form of free speech — stands for. But while this sort of thing is going on, you'd better just double-check those privacy settings, and cancel that account on that swinging site you signed up for "just to take a look."