I have read in the privacy community that facial recognition done in public places is considered problematic. Not knowing what is considered the crux of the matter, I have to ask about some facial and behavioral recognition use cases here, and whether they are a problem or not
- Digital signage on roads, cameras in stores, etc. read pedestrians’ faces, movements, etc. and infer attributes for marketing purposes (sometimes the inferred attributes are stored as is, sometimes they are stored as statistics and the attributes themselves are removed)
- Public transit agencies can share police databases to identify and track individuals with arrest records
- Public agencies use facial and behavioral recognition to determine and track suspicious persons. The information read is stored.
With behavioral identifiers, it is possible to change the type, quality, quantity, or values of data collected about you going forward (and thus your ability to be tracked) without too much of a problem. You can buy different devices or trade your existing ones, avoid specific websites/apps, use privacy friendly services, use services that protect you, change/delete your information, provide false information, etc. Granted, there may be limitations and certaintly many inconveniences in doing so but it is possible.
When it comes to biometric identifiers (eg - retina, fingerprint, face, dental, DNA, etc.) or even pseudo-biometric identifiers (eg - gait/stride, handwriting, typing cadence, etc.), it is much more difficult, if not impossible, to change the type, quality, quantity, or values of data collected about you. You may be able to change all or some of these identifiers, but at a much larger financial cost, inconvenience, physical pain, etc. and you may only be able to change some of those identifiers a finite number of times. How do you change your DNA (or prevent companies from obtaining it with certainty) once insurance companies have it and start denying you policies due to certain genetic dispositions? How do you change your fingerprints once a company who requires that information to prevent abuse of their service is hacked? How do you change your facial construct after a company like ClearviewAI has it from your presence on social media?
You may be OK with a certain company having your data right now for a particular purpose (eg - biometric lab for medical blood tests), but you may not be comfortable with that same company having that data years from now (eg - same biometric lab keeping your data to help identify medical trends for you, but you no longer want them to have it because you switched labs), using it for purposes unbeknownst to you (eg - same biometric lab uses your data for medical research), or any of the entities that have received that data via sharing, selling, hacking, negligence, etc. Once your data is made available or placed in the wrong hands, you have no idea how your data will be handled, used, or abused nor for how long your data will be stored. That’s especially true for data that cannot easily be changed, like biometric data.
Also, there are several examples of exploitation, misidentification, and other poorly implemented versions of facial recognition. People have been misidentified as someone committing a crime and incorrectly arrested or even placed in jail. People have been incorrectly targeted by stores as a person likely to shoplift due to their skin tone. It’s also possible for fraud or identity theft to occur when this technology is handled poorly.
Despite all this, many retail companies are including this type of technology and it will likely continue to be adopted unless there is some legal or regulatory rrestriction Additionally, it’s never been easier for people to feed the facial recognition databases (eg - doorbell cameras, home security cameras, cellphones, drones, etc.). The dystopian scenario behind this is that you won’t be able to leave your home without being tracked between consumer devices in residential areaa, government devices in public places, and corporate devices in retail/office spaces without being tracked. The scary thought is that we may be closer to that scenario than many people realize.
Thank you for your detailed response.
In my country, biometric and pseudo-biometric data acquisition is widely used by companies and even the government for crime prevention (and of course for marketing purposes). Where do you think there should be a trade-off in privacy vs. crime prevention (benefit to society as a whole)?
Should we have cameras? Absolutely! They’re great at providing evidence of a crime and catching the perpetrator. They can be used in ways that doesn’t egregiously invade everyone’s privacy that happens to walk by a camera. Should there be automated human/object detection? Sure! It could help reduce storage costs of video and maybe even enable more cameras to be used as a result. Should the videos have automated individual identification/facial recognition? I’m torn here. Having it could enable certain features. For example, a bank could be setup to only allow the vault to be opened for certain individuals or a person could setup an automation to automatically unlock their front door when they walk up. However, I do not think that massive databases of people/faces should be built, collected, shared, etc. There are several other ways to prevent and deter crime that do not require the invasion of everyone’s privacy. If the concern is theft, you could place the items in a cage, behind a counter, in anti-theft boxes, hire additional staff, etc.
Out of all the other reasons listed, I don’t see how most of them benefit society. What good does facial recognition provide on tracking pedestrians? Why is it needed? Couldn’t whatever benefit you think is occurring happen with cameras that only detect that a person is on screen? Why would you want facial recognition for marketing purposes? What benefit does it provide you or society that the dozens of other ways marketing companies track their target audience isn’t already doing? Also, why do people who aren’t part of a target audience need to be included in a company’s mass collection of faces? Why does a beard care company need to gather women’s faces? Why does a tampon company need to gather men’s faces? Why do we need another way to track people taking public transit that couldn’t already be done with existing methods?
The only reason that I believe you’ve listed that is potentially valid is public agencies tracking fugitives. If you’re wrongly committed of something though, what do you do? You can’t go in public, you can’t take public transit, you can’t go to the store. Do you just turn yourself in and hope the agencies find the true perpetrator or hope you can afford a good enough lawyer to get your case dismissed? Also, what percentage of the population requires this? I assume its likely less than 10% and that may be generous. So that means that the vast majority of the population do not deserve to have their privacy protected and must have their faces collected, sold, shared, etc. because a small percentage of the population may be fugitives.
Crime has occurred before this privacy invasive technology existed and it will continue to occur with it. It’s unreasonable that everyone’s privacy must be abused in the name if safety when it will likely not even make a major change to crime prevention.
I agree with your statement. Then, if I dare to respond with my country and citizen stance…
In fact, besides facial recognition, a complex set of technologies is used. This allows us to infer information such as where, who, and what they did. Such systems are often implemented from a crime prevention perspective, but they are also useful for large scale collection of statistical data. The results of the analysis are fed back to us, and many people are happy with the increased convenience.
It may be a hidden target, or it may be useful in another business of that company or government. In fact, some services have been launched as a result of such data analysis, with some success.
The behavior of fleeing because of a false accusation is likely to result in another charge or more accusations than being arrested. You should either turn yourself in or be arrested and fight it out in court. Others argue that these systems are also useful in reducing false convictions.
Also, many citizens support such a system, saying that it is “for the safety of society” and “for the arrest of criminals.”
Because of the laws governing the handling of personal data, many citizens trust that the information obtained by companies and governments is properly managed and used for the right things, and the prevailing view is that sacrificing some privacy for security and convenience is unavoidable.
My privacy is more valuable than your safety.
Those same people wanting safety are often the ones wanting criminals to have lighter sentences.
It was introduced not to lighten the sentence, but to minimize the harm suffered by the victim. Support is strong, considering what would happen if I were to be victimized tomorrow and what would happen to the victim.
Why do you need to know who is doing what? Wouldn’t it be nearly or just as beneficial knowing someone is doing something? You could use person detection (but not identification) for this. If you were trying to see how many people used a park, person recognition could help solve that.
Abusing people’s privacy is a large price to pay for new business. Once a population’s privacy is lost in certain areas, it can be really hard to get it back. Typically we outlaw businesses where people are abused, whether its their rights/freedoms, health/wellbeing, safety, etc. I’m not sure why privacy should be any different (I know privacy isn’t a right in all countries though).
I agree. However, I don’t think it’s clear or obvious what they are fully agreeing to. I imagine if all information you have on someone, the insights drawn from that information, and possibly even how they information is/could be abused, I think their stance on the topic may change. To many people, their privacy is something that they don’t realize they want and need until its clear to them that it’s been invaded. Since this these facial recognition programs are done in ways that don’t prevent people from doing what they want to do (eg - go to the store, go to the park, etc.), are unobtrusive (eg - the cameras are silent, placed in corners, relatively small, etc.), and the processing is done behind closed doors, people don’t really care or understand what is happening. They may not know how many cameras have captured their activity or even where the cameras are that may capture their activity.
If we replaced wall mounted cameras with people operating a camera and an individual camera operator follow each person around in areas that are using facial recognition, do you think people would be as supportive of such programs? I don’t think that the additional cost of cameras/operstors or crowd due to camera operators will be people’s main complaint.
Even if we just sent a letter to each person identified by facial recognition cameras with the information that was gathered (eg - where they were seen, how long they were seen at each spot, who they were seen with, what they may have been doing, etc.) plus any additional information that is tracked (rg - you were seen at this location XX times in the past YY days/weeks/months, you are typically at this place with ZZ person(s), you typically spend XX time there, etc.), then I doubt people would be so apathetic to their privacy being invaded.
For example, by estimating who visited the location, to do what, and how many times, they are able to determine if there is any suspicious behavior (although the criteria are unknown). This is, of course, also used for marketing purposes, as in how many times a customer with what attributes has used the service.
If there is an alert that there is suspicious activity, the police or employees will monitor, track or question the person in question, and the system will mark that suspicious person as one of the people on the suspicious person list in the future.
Also, there are many instances where companies and governments have built facial recognition systems, saying that they are doing person detection. It is difficult for the public to be sure that companies and governments have not built and abused such systems.
Yes, privacy is a very low priority right in my country. For this reason, people often adopt the idea of sacrificing privacy to protect the rights of citizens or to make them healthier and happier.
Your response raises a lot of questions and potential red flags.
Who is developing the algorithms that determines suspicious behavior? Who is reviewing those algorithms to determine if any bias is in the algorithms, regardless of whether the bias is intentional or not? Are changes to the algorithms retroactive (eg - does a person who stood in a place for 11 mins last month get an infraction when the algorithm is updated from <=15 mins in that location is OK to >=10 mins is not OK)? How long are the infractions held for? Are immediate actions taken when an infraction occurs (eg - guard dogs are released when a criminal gets too close to an expensive item)? Are people made aware of any infractions they may have or receive? Are people made aware of what causes any/all infractions? Are people made aware that they’re being recorded or the level of recording (eg - facial recognition vs person detection)?
The point I’m trying to make here is that there are so many things to consider to make sure this is done correctly. I don’t think we even understand all of the scenarios to think through. Its one thing to do things wrong on a small scale, but when you’re talking about surveiling an entire city or country, then this needs to be done extremely thoughtfully and cautiously.
This goes back to the point I was making previously - it’s really hard to leave gihr home and not be recorded already. You’ll likely never know where you’re being recorded, who is recording you, or what will be done with your recordings. We’re losing the little privacy we do have left outside our homes and there’s not much we can do about it.
Since most of these digital privacy invasive technologies are innocuous and people are told of the benefits without understanding the drawbacks, its easy to think this technology is good. However, as people start to understand the scale and capability of these technologies, they will likely become less happy about them. They may get frustrated at not only themselves for not opposing the technologies originally, but at the government, retail store, tech companies, etc. for promoting the technologies and exploiting them.
Also, being monitored and tracked all the time is not a healthy thing. People behave and think differently when they’re watched. People need alone time. People need privacy. Having it allows people to explore, be curious, experiment, learn, and so much more. Having facial recognition cameras may not seem too repressive, but when it starts to change their behavior because they don’t want to be identified at a certain park/store/corner/etc., then that mental/emotional barrier to entry is just as hard to overcome as a physical wall/gate.
Only companies and developers would know the details. Of course they are auditing, and if you ask, you will probably get some kind of response. On the other hand, they may not tell you for security reasons.
That is put up as a poster, but some people may not know. If you want to use that service, you have no choice but to accept it.
If we’re going to lose them sooner or later, wouldn’t it be better to make use of them in order to achieve a more prosperous society?
The state also justifies the arbitrary harvesting and storage of biometric information by claiming in court that "it is the natural authority of the state to store information about its citizens. If this is followed, there is already no privacy at all in the public sphere. And large corporations will follow the state’s argument and say that it is their natural right to supplement their customers’ information.
I guess my country is trying to figure out how to enrich our society in the future by acknowledging that privacy does not exist. And shouldn’t we?
This is a statement I heard from an acquaintance of mine, but I’d like to ask it here as well:
Do you know for sure the algorithms are audited? Has a verified 3rd party publicly reported their findings? A government or company saying “we were audited - trust us” holds as much credibility as them saying “we’re the best government/company in the world - trust us”. The purpose of having auditors is to provide assurance to a group of people (eg - shareholders, potential investors, citizens, etc.) without publicly exposing trade secrets, security details, or other sensitive data. If a government/company takes transparency and ethics among other things seriously, they would regularly have their algorithms audited and the findings would be made available ins relevant way.
No - absolutely not. If we had this mindset for every social issue, we would never have progressed as a society. If the government were to come and take each citizens possessions every few weeks, would you just say “well, I trust they’re making better use of my possessions than I would have. It’s surely going to benefit my society better if they have it than if I did. I might as well might head down to the townhall each week and give them whatever I acquired that week to make it easier for them.” I’m sure it may sound a bit silly or hyperbolic, so maybe think of a people who feel their taxes/garnishments/tithes/etc. may be a little too high - same principal. Now re-read that but replace “possessions” with “privacy” because that is exactly what we’re doing. People may be OK with things because they don’t understand what is going on. Following this analogy, as soon as people find out that their possessions are being misused, people are going to be very upset. Likewise, once they realize their privacy is being abused, they’ll be very upset. It may not be this year or this decade, but continuing down the “well privacy doesn’t exist, so let’s keep abusing people’s privacy further” path will eventually lead to a very upset population.
Just because the government is doing/has been doing this doesn’t make it right. However, your government must have restrictions to this, right? For example, they’re likely not doing annual home inspections to look for signs oof"resistance". If they are, do you genuinely think people are OK with this? If they’re not doing that, do you think people would openly and willingly allow and encourage that behavior?
Same for the Disney comment. Find me a privacy rights activist who is OK with Disney checking fingerprint, but not small businesses doing something similar. I imagine you’ll have a very hard time doing so. I hope this is outlawed and/or Disney (and any other company that unnecessarily collects biometric information) stops this behavior because Disney most certainly does NOT need anyone’s fongerpirnts in order to operate their theme park.
deleted by creator
I was pissed when TX fingerprinted me just for getting a driver’s license.
Thank god for hoodies and mirrored sun-glasses. men should all start wearing their hair a bit longer to hide the side views of their face.