The "right to be forgotten" has created a right old headache for Google and for publishers alike. As a result of a European court ruling, people can request results relating to them to be removed from Google's searches – even if the original article is true, fair and accurate – if it is "outdated" or "irrelevant". Google says it has received at least 70,000 such requests.

So far, six Guardian articles, at least one BBC article and several Mail Online pieces have been affected by Google's implementation of the ruling, we learned on Thursday, as the publishers received automated notifications that their work had been partially delisted. There has been no small degree of confusion as to what this actually means. Firstly, no web page disappears from the internet as a result of this: the articles are still present, but may become much harder to find because Google removes the offending article from search results for the complainant's name.

So to give a hypothetical example, if the Guardian ran a story about a John Doe running an illegal kitten farm in Surrey, and years later he complained to Google, the article may be removed from search results for "John Doe", but could still appear in results for "illegal kitten sanctuary Surrey".

This can have some strange effects: the "disappeared" BBC article related to Stan O'Neal, former boss of Merrill Lynch. However, it seems O'Neal was not the person who complained about the article, which still appears in results for his name; instead the complaint came from someone who commented below the piece.

This can still be a serious challenge to freedom of expression. "Outdated" and "irrelevant" are broad terms that lie in the eye of the beholder, especially as the court neglected to give specific definitions of either in its ruling. The Guardian or Daily Mail might regard a 2005 article about legal tax avoidance as timely and relevant if its subject is still doing business. The subject of the piece is likely to feel differently. The ruling makes Google decide.

This gives Google a huge role in policing the internet, and one it really doesn't seem to want: if it tries to resist too many (or any) removals, it could easily end up mired in numerous cases with Europe's information commissioners, taking up huge time and expense for the company. Similarly, though, it doesn't want to be left with thousands of daily requests for manual tweaks of search results.

It's convenient, then, that it's found a way to get the media to kick up the fuss for it: there are very few news organisations in the world who are happy to hear their output is being stifled. A few automated messages later, the story is back in the headlines – and Google is likely to be happy about that.

Still, just because Google might like changes to the ruling, or how it would work, doesn't mean the rest of us might not also benefit. At stake is the internet's effectiveness as a research tool: do we think the internet is better if people are free to wipe the slate clean, even if they really did everything that was written, or not? Who should get to decide what stays and what goes? The time to decide is drawing very close.