I think the "right to be forgotten" should be replaced with the "right to not have old information considered", which is a much better compromise between the public interest and privacy. In your example, a person false accused of a crime wouldn't have the record expunged, but the record would be heavily tagged with notices that it was a false accusation, and it would be illegal to use that information in considering him for a job and it would be a severe breach of social norms to use that information turn him down for a relationship.

Considering old information, whether it's relevant or not, is something human psychology does automatically and inexorably. So when you talk about a "right to not have old information considered", you might as well speak of your right to jump off a cliff and soar through the air instead of hitting the ground. These European laws are doing their best to account for the unfair tendencies of human nature, while your counter-proposal isn't really an improvement over the situation we had before.

> "... but the record would be heavily tagged with notices that it was a false accusation ..."

There's been some literature lately about people trusting their doctors/lawyers/etc. more when that person reveals that they have a conflict of interest. A similar backfire effect could well occur here.

> " ... it would be a severe breach of social norms to use that information turn him down for a relationship ..."

Oh good. Approximately when can we expect the social norms to change? And are you accounting for the fact that it may happen never?

/Noticing/ old information is automatic, but assigning valence to it and having it influence a decision to a significant extent isn't. I might not be able to notice that a person is left-handed, but that doesn't mean I will favor or disfavor them because of it.

> There's been some literature lately about people trusting their doctors/lawyers/etc. more when that person reveals that they have a conflict of interest. A similar backfire effect could well occur here.

A file stating that an accusation was false isn't admitting doubts or uncertainty, though, and it's that admission that leads people to trust advisors more when they reveal conflicts of interest. That file is flatly and loudly stating that the accusation was false.

> Oh good. Approximately when can we expect the social norms to change? And are you accounting for the fact that it may happen never?

Oh good. Approximately when can we expect law to change? And are you accounting for the fact that it may happen never?

That's interesting and would help a bit, but its probably not going to help the man falsely accused of murder very much.

In the relationship case its basically pretty much impossible to legislate norms of behaviour in any reasonable period of time (it would be multi-generational if possible at all). For the job applications we all know hiring managers discriminate even when its against the law due to either their explicit or implicit biases. They can find very reasonable and legal explanations for rejection that they probably even sincerely believe themselves.

I'm taking the long view of things, so I'm less interested in one particular man falsely accused of murder than falsely accused people in general. Multiple generations is indeed a reasonable period of time. Sure, social norms might change slowly, but that slow change is often the only way to fix the problem without breaking something else to an unacceptable extent.

That being said, there is role for legislation--for example, there is a lot of research into implicit and explicit biases in human minds, and there could be more public funding on how to reverse these biases, and implement them as a standard part of the curriculum in public schools.

> removing a factually true report that is just two years old seems a bit extreme

It might also be the case that old articles (say, of 25+ years) should be put back into digital archives. In such a long time the persons involved are no longer affected, but the public interest might exist, for historical archival reasons. So there would be an "invisibility interval" from year 2.5 to year 25.

Totally agree - there should be some friction. I'd say this should put more pressure on search engines, but not on the content archives, so that potentially damaging information could be found only by precise search of the events in the specific period and location regarding the person of interest, not just by name. However, direct search in archive must remain unrestricted.

Probably, there should be a legal mechanism similar to US FoIA regarding the personal information collected by non-governmental entities, in accordance to which it could be possible to request the dossier collected by a company on you and use it as a protection from discrimination (as in your first example).

It could be something as simple as you need to pay a fee for a search query in order to get hold of filtered results, but with no indication that there will be any extra results to find.

eg: Search 'John Smith Liverpool UK Murder' and you get nothing interesting and no indication that anything has been filtered. However there is an option that is always available no matter the query to buy this search query for say $1-$50. If you pay up you get the filtered results, which may happen to show that John Smith was once accused of murder many years ago. If you are an investigator, researcher, or journalist then there should be a flat single fee or free access to filtered results.

>E.g: Man falsely but not maliciously accused of rape or murder in press reports and later being exonerated in court without accompanying press report detailing the exoneration, and now finds it very difficult to get a job or start new relationships. Or a business man who filled for bankruptcy 30 years ago who finds search engines return news reports of his bankruptcy as the first and only return on himself and thus makes finding business partners or credit near impossible.

I'm not trying to blame the victims here, but if you're in these unenviable situations, wouldn't it be possible to avoid these problems by changing your name? If Google searches for "John Doe" (obviously, a more unique name than that) return press reports about you being tried for rape, changing your name to "Jim Smith" is going to eliminate that problem handily. A potential dating partner would have to do a significant background search on you to find out you changed your name, and then look up that one.

I mean, it sucks that you'd have to go to the trouble of dealing with the courts to change your name, and also have to lose your name which you might have an emotional attachment to for understandable reasons, but for infrequent cases like this, this seems like a much easier thing to do than to try to get data warehouses around the world to stop reporting past stuff about you in response to queries.