Amazon Pushes Facial Recognition to Police, Prompting Outcry Over Surveillance (nytimes.com)

Nick Wingfield, reporting for The New York Times: In late 2016, Amazon introduced a new online service that could help identify faces and other objects in images, offering it to anyone at a low cost through its giant cloud computing division, Amazon Web Services. Not long after, it began pitching the technology to law enforcement agencies, saying the program could aid criminal investigations by recognizing suspects in photos and videos. It used a couple of early customers, like the Orlando Police Department in Florida and the Washington County Sheriff's Office in Oregon, to encourage other officials to sign up.

But now that aggressive push is putting the giant tech company at the center of an increasingly heated debate around the role of facial recognition in law enforcement. Fans of the technology see a powerful new tool for catching criminals, but detractors see an instrument of mass surveillance. On Tuesday, the American Civil Liberties Union led a group of more than two dozen civil rights organizations that asked Amazon to stop selling its image recognition system, called Rekognition, to law enforcement. The group says that the police could use it to track protesters or others whom authorities deem suspicious, rather than limiting it to people committing crimes.

Comments

Should law infocement be hard?

[jellomizer]

I get that Law Enforcement in general is trying to help the public, and are trying to find tools to make their job easier and more efficient. However in order to protect our freedoms law enforcement needs to be hard work, even if it means our lives are measurable less safe.

We cannot have Safety and Freedom. For increase safety there is a trade-off in freedom. While there may be some rules that will increase safety by a factor of ten and reduce freedom by one tenth, and may be considered a fair trade off, there are other things that may give us marginal safety benefits with a large hit to our freedom.

Law Enforcement professionals work with the scum of the earth all the time, this is affecting their judgement, and their job is to keep people safe. So I do not fault them for wanting more tools to make their job easier and more effective. However we as citizens need to stand up and say. "We thank you for the effort and your hard work. But we can't let your job be easier at a high costs of our freedoms"

Re:Should law infocement be hard?

[imrahilj]

I think this is the way to think about things. The problem is that once a freedom decreasing measure is tried, it brings along with it a bureaucracy that will argue for it's continued existence, regardless of effectiveness. The TSA brought us a decrease in freedom, without an observable increase in security. Now that it exists, however, any discussion of its failings will not lead to a discussion about dismantling the TSA, it will only lead to discussions about how it must use more resources and be more intrusive. The ratchet swings only one way.

Re:Should law infocement be hard?

[b0s0z0ku]

Nah, most dumb cops in the US just see their careers and an opportunity to control people. They don't even think enough to think "better society."

Re:Should law infocement be hard?

[Bob the Super Hamste]

Most cops went into the job because ... , washed out of the military ...

I seem to remember reading or hearing about that cops who were former military were less likely to shoot someone and generally did a better job. Yup it appears that I did read that so I would prefer to have more ex-military cops given the available evidence.

Well ...

[cascadingstylesheet]

Well ... if Amazon is selling this to whoever, then I'm sure some "protesters" will also use this to face ID people at the "wrong" political rallies, etc.

Since "protesters" love to get people fired, blacklisted, harassed at home, etc. that should be fun.

It's the tech genie. You can force yourself to put it back in the bottle, maybe, but you can't really force everyone to do so.

Re:Well ...

[Virtucon]

Oh I'm sure there are already organizations who've signed up but presumably only state/local gov't agencies would have access to the camera networks. That doesn't preclude the use of a few well placed go-pros at protests that could also serve as a data source as well.

where is the line?

[supernova87a]

Ok, I know there is somehow a fundamental difference between a computer system with unlimited memory and processing power, versus a person who has really good memory.

But entertain me on this thought experiment. Why is having a police force use such a system so different from if they had on their payroll someone who was really really good at remembering faces? Or someone who knew everyone in town?

At what point is an automated / faster system an unreasonable infringement of your rights compared to what each of us can do to some degree? Is it the natural size (200-300 people?) of our memory and human facial recognition that sets the limit on what is an invasion of privacy or not? Where is the line? What is different about using this system compared to a police officer asking everyone he/she can find whether they know person X?

I find the definition of reasonable privacy difficult to nail down.

Re:where is the line?

[DontBeAMoran]

NO! NOOOOOOOOO!!!
I will not sacrifice our privacy rights. We've made too many compromises already, too many retreats. They mine our personal data, and we fall back. They assimilate entire databases, and we fall back. Not again! The line must be drawn here! This far, no further! And I will make them PAY for what they've done!

Re:where is the line?

[Anonymous Coward]

You're missing the point. No one actually really cares about them scanning faces.

What people care about is other people then using that data to do bad things, stalking, harassment, silencing political opponents, blackmail etc. All of which are things that have repeatedly happened with state and private surveillance. The real privacy issue is that you can't trust others to actually be responsible and hold themselves correctly accountable to even entertain the idea of giving people this power.

Re:where is the line?

[Kiuas]

But entertain me on this thought experiment. Why is having a police force use such a system so different from if they had on their payroll someone who was really really good at remembering faces? Or someone who knew everyone in town?

Okay I will try my best. I Just recently I warthed the documentary about automated surveillance systems sold to and used by the police both in the US and here in Europe called Pre-crime after the concept from Minority Report. The doc itself was alright, not the best, I would have liked more details and practical examples but it gives some insights into the kinds of systems currently in use and in development. Essentially these kinds of systems can be divided into 2 main categories:

1) The kinds that use open and public data about crimes and do not involve personal factors like face-recognition. These systems are meant to help the police to figure out which areas should be patrolled in more, and where the crime is currently clustering.
2) The kinds using both private and public information. These take personal information from different sources, including but not limited to social networks, financial systems and for example state services like social security, past arrests etc and 'score' individuals to try to give a number to them indicating how likely it is that said individual will either perpetrate or be the victim of a crime. Such systems are currently in use at least in Chicago, parts of the UK and parts of China.

Okay so, take your example and assume you have an officer on duty with supreme photographic memory that's capable of remembering a lot of faces. He sees someone on the street that he thinks is a fugitive or a suspect and they pursue him. If something goes wrong and the guy's injured or killed and turns out to be a different individual, the cop can be held accountable legally. However what's happening with these type 2 systems and what the documentary describes is that it takes and unknown number of variables, and then uses an unknown algorithm to come up with a score that's supposed to tell the officers who is a danger or who might be in danger. In the doc they mention one of the guys who's in the 'heat list' of around 400 individuals in Chicago that has a 'high threat score', not because of his own past criminal history but because of who some of his friends are. Now, with such a system if you happen to end up on such a list there really is no way currently to get of it. This means whenever these guys are walking out in public if the police pull up information on them from the system they're far more likely to be stopped.

Now imagine someone like him, who doesn't have an extensive criminal history but is on the list due to his personal connections gets into an interaction with the police. Do you think the cops will treat such a guy in the same way as they would any other normal individual? Even if the guy's doing nothing wrong a the moment, the system is there to remind them that this is a 'high risk' target. Do you think that increases or decreases the chances that officers will resort to use of force because their algorithm is telling them this guy might be dangerous?

That's the problem core of the problem. It's not the same as some guy seeing a face he remembers from a past case on the street or knows to be a gang-member. These estimations seem objective, but the fact is we're still dealing with probabilities at best. What the scores generated by such systems are telling you is that there's an N percent change according to them that this guy might at some point commit a crime. But there's essentially no way of telling whether those probabilities are accurate for a given individual, yet the police trust the score gioven by the system. One example mentioned in the doc was that a woman had been tweeting about a card game by the name of 'Rage' often, which the system interpreted as a sign of aggressive behaviour which raises her score. But the cops on th

Re:What "outcry"?

[kilfarsnar]

I am American and many people agree that the vast majority of the public SUPPORT this kind of law and order in this country and you libtards are the minority, AGAIN. This is why we support who we support even though it makes you weak snowflakes so angry, the people who will make this country great are the ones who oppose the city thugs who terrorize our daughters and cuckold our sons.

See, this is why the rest of society thinks you're dumb. You have not thought this through, and it shows. You only support this because you think it will be used on those you deem undesirable, whether that be people convicted or suspected of crimes, or people in this country illegally, or "libtards". But there is nothing stopping it from being used on you and your fellow travelers; at which point you will cry about your rights and how unfair it is.

What if this service were used to keep track of NRA members, or people in a citizens militia? What if it were used by the FBI to watch for people wearing MAGA hats? Would you still support it? Or would you recognize it as the affront to your rights that it is? Whether you like it or not, when one group loses its rights, it makes it easier to take them from another group. You should fight for the rights of all people, not just those you like or agree with. Because that's how you protect your own.

Re:What "outcry"?

[gweihir]

Indeed. The reason fascist states can still be constructed is that there is a large population of morally-challenged morons that a) are willing to apply any amount of violence to anybody that is not like them and b) that have no clue that they are just a bit later in the chain of victims.