Hacker News new | past | comments | ask | show | jobs | submit login

What qualifies as part of the facial recognition chain? Does source video count?

Because I could see an interpretation in which people filming police activities and the people they arrest as no longer being allowed.

What about that?




If they are running automatic facial recognition on the video then yes, it's banned.


The social media loophole seems ripe for exploit. As long as the destination of the content goes to a social platform it's all free game.

I wonder what they will consider a social platform as people develop new systems to get around this law. Cuz logic and unknown reasons.


A lot of the comments here seem conflicting, to say the least. The article has:

> use by private entities in public accommodations

Which still doesn't fully explain it for me. In an HN comment someone linked what appears to be the ordinance in question: (https://static1.squarespace.com/static/5967c18bff7c50a0244ff...)

That says (§34.10.020 D) that a place of public accommodation is a "place or service" offering various things. I'm not quite sure how to interpret the wording "advantages, facilities, or privileges" though. In addition to public businesses, I'm assuming that would include city parks (they seem like facilities), but what about roads and sidewalks? I guess those offer "advantages" and "privileges", at least?

But regarding the context of the earlier comments in this chain, it seems like intent would be the key factor. Recording video with the intent to run facial recognition against it, whether it was to be done by you or someone else, would presumably qualify as a violation.

So what about recording large quantities of video without intending to run facial recognition, for example a dash cam, but then later running it against a specific segment in response to an incident?


Recording video with the intent to run facial recognition against it, whether it was to be done by you or someone else, would presumably qualify as a violation.

It seems to me quite clear that the act of running facial recognition is the problem.


I'm not clear on your meaning?

In context, an earlier comment asked about the legality of filming police activities under this ordinance. I was pointing out that intent typically matters to the courts.


What's unclear? The law is written about the act of running facial recognition. Videoing isn't banned, writing facial recognition software isn't banned etc.

Sure - intent matters - but it's the act of running the facial recognition software that is banned.

If you start videoing police and were stopped by police and the police could show you were intending to run it through facial recognition then the court might find them stopping you was legal based on this.

But if you are filming police and live broadcasting it on Twitter or something then it's pretty unlikely a court would find stopping you was legal under this law. But if you streamed it to someone else having previously organised for them to run recognition then stopping might be legal.


Yeah that is a silly loophole. If it's dangerous enough to ban ( and it is) how can social media be a justifiable exception. Unless they implicitly trust the people that run social media companies for some reason?


My guess is that they lack the jurisdiction to regulate how (for example) Facebook processes images on their platform. Rather, the ordinance only applies to entities while they are within the city. So without the social media exception, it would likely be illegal for any individual within city limits to take a picture with the intent to upload it to Facebook. That would make for some interesting fallout to say the least.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: