Katy Guest: I used to like Facebook. How did it all go wrong?

Reporting six months ago on the announcement of an initial public offering by Facebook, this paper sounded a warning to the social networking site. "An increasing number [of users] are likely to feel bruised as they are confronted with the bitter truth that they are mere fodder for a machine that means business," wrote our consumer correspondent. A marketing expert added: "It takes clever leadership and in-depth understanding of where you can introduce business elements without destroying your value for users."

Last week, stock in the company slumped to a new low. Many investors jumped at the first opportunity to offload their shares, reducing the value of the company to £34bn, from £104bn at its debut. To some who use Facebook, it was just desserts.

In many ways, the rise and potential fall of Facebook can be seen as a metaphor for the internet. It was invented in 2004 by a team of college students to fulfil a need that nobody knew they had, and quickly became one of the biggest companies in the world. It went "cash flow positive" in September 2009 with an annual advertising revenue of hundreds of millions of dollars. But although the internet has been around for longer than Facebook's founder, Mark Zuckerberg, still nobody really knows how to use it. Or to "monetise" it, as the reports in the past week's business pages put it. Many of the best things on the internet – Wikipedia, Google, online newspapers – are free. Free to use, that is; not free to produce.

There are different ways around this, but each of them is a compromise. It's easier for websites with an actual product to sell. Others, such as Wikipedia, ask for donations. Some newspaper websites erect "paywalls" so that readers can pay for the journalism that fills them, but it's a bold move when there is the BBC, giving away news for nothing. Others hope that advertising will cover the costs. Facebook's costs, for hardware, data storage, energy and staffing, must be in the hundreds of millions. Its compromise is that its users are its product, and it sells them to advertisers in order to survive. It's an uncomfortable arrangement.

As of June 2012, Facebook has over 995 million active users, and I'd bet that about 954 million of them didn't read the "User Agreement" before signing it. (Its "Terms & Conditions" print out at six pages of small-print A4, not including 10 further large documents such as "Data Use Policy" and "Advertising Guidelines".) In which case, we really only have ourselves to blame. When I signed up, about five years ago, Facebook seemed a fairly benign organisation designed to keep me in touch with my cousins. But recently, I've found it increasingly sinister, and I'm not alone.

First, there was an issue with my privacy settings, which I had painstakingly set to "friends only" – meaning that only people I had actively agreed to be friends with on Facebook could see anything I posted there. One day, I found that they were all set to "public", meaning that anyone with a computer could now see my holiday snaps. "Nothing changed to public without you choosing to set it that way," Facebook told me. "There are misconceptions about privacy settings but the changes last year were about simplifying the settings and adding in inline controls as you post." I was confused about the difference between "change" and "simplification".

Then I objected to some of the targeted advertising. Not the often interesting ads directed at me because I live in west London, recently got engaged, or occasionally write status updates about dark chocolate Magnums. But adverts directed at me just because I am female. Telling me to lose three stone in a month. I find them sexist, offensive, and unbelievable, and so I deleted them and asked for them to stop. "We will take this feedback into account as we continue to improve our advertising systems for both advertisers and people that use Facebook," said Facebook, months after I complained that the ads hadn't stopped – won't stop. "Please keep in mind, however, that if you choose to provide feedback for an ad, this may not immediately affect how often you see this or similar ads." Still, they come.

More recently, though, something else has started to smell fishy. Facebook enables its users to "like" things that they see on the site. By clicking a button under a friend's photo, or a dark chocolate Magnum, say, you can let your friends know that you approve of it. It's a silly and easy new way of communicating. But according to Facebook, I "like" Amazon Kindle UK. This is annoying, because I didn't, and I don't. And, at the time when I supposedly clicked "like", I was fast asleep in bed. It's even more annoying because I am the literary editor of a newspaper, and some people mind about my opinion of Amazon Kindle UK. I deleted my "like" and reported it to Facebook, as did dozens of my Facebook friends, and their friends, who had also somehow been made to "like" Amazon Kindle UK when they didn't. But it kept coming back. Eventually, I contacted Facebook's press office and asked them to investigate.

In one of the first emails in a conversation thatcontinued over a three-week period, they sent me detailed instructions on how to delete the "like" at its source, which I did. In one of their last emails, they told me: "As the 'like' has been deleted, unfortunately there is not enough information to investigate any further." They were also unable to investigate three other fake "likes" that a friend had left undeleted solely to give Facebook a chance to enquire further. ("I'd be surprised," said a contact at the anti-virus software people Sophos, "if they couldn't look through their archives of data and see what happened when on your account.") Facebook suggested that the "like" could have been put there by "malware" on my computer, though I've told them that my computer has been checked by IT experts, and no malware or viruses have been found. It added: "We investigate all reports of spam quickly and take action." Next, I explained the problem to Amazon, whose PRs said that they understood my deadline for this article, but no one returned my calls.

Recently, the BBC's brilliant "Virtual Bagel" sting discovered millions of fake Facebook accounts, which existed apparently only to "like" commercial pages. Facebook admitted that 83 million, or 8.7 per cent, of its accounts were illegitimate. But it's one thing for a fake person to "like" a real product, and another for a real person to have "like" wrongly attributed to them. If a publisher were to write on a cover "Katy Guest likes this book", when I didn't, there would be trouble. I find it hard to see a difference here, but I'm still baffled as to who might be doing this, and why.

The Advertising Standards Authority says: "Endorsements or testimonials used by advertisers must be genuine, whether they appear in a Facebook-sponsored story or in any other advertising medium, and advertisers must hold evidence to that effect. We have yet to receive complaints about Facebook 'likes' so to date there is no test case in this area. We stand ready to assess any complaints which suggest that false endorsements are being used in ads in social media or elsewhere." The Information Commissioner's Office (ICO) tells me that "any social media personal page is personal data, so would fall under the Data Protection Act". Facebook claims it takes security very seriously and constantly develops new tools to detect fraudulent activity. But if it can't tell me what went wrong here, a complaint to the ICO might have to be my next step.

I know that I'm not the only person who finds all this sinister, because I've garnered opinions on Facebook. Yes, I'm still using the site. For now. But the internet being what it is, it can't be long until a bunch of students finds a way to provide something we're now all crying out for: Facebook, as it used to be, five years ago. I'd like that. I might even "like" it. Facebook's shareholders, on the other hand, might not.