Monday, July 7, 2014

My Insolent, and Fully Rebuked, View on Surveillance

I previously posted this on another blog under the title: Revisiting Surveillance with the ACLU as my Guide. With minor edits, and the addition of a video, I believe it belongs here in this blog.

In an American Civil Liberties Union (ACLU) position paper on surveillance cameras used by police (known in the title in a slightly less threatening phrase as "Police Body-Mounted Cameras") the ACLU tells it all in the subtitle: "With the Right Polices in Place, [such surveillance is] a Win for All." link

It is sort of startling to me because I thought I was living in a world where all surveillance was deemed inherently evil by everyone around me. This I gathered by the reception I got when arguing the proposition here on the Internet that I felt safer when someone was watching.

Here is the final conclusion of the ACLU position paper from 2013:

Although fitting police forces with cameras will generate an enormous amount of video footage and raises many tricky issues, if the recording, retention, access, use, and technology policies that we outline above are followed, very little of that footage will ever be viewed or retained, and at the same time those cameras will provide an important protection against police abuse. We will be monitoring the impact of cameras closely, and if good policies and practices do not become standard, or the technology has negative side effects we have failed to anticipate, we will have to reevaluate our position on police body cameras.

This has been obvious to me, and I am delighted the ACLU has taken my side in some instances albeit with huge caveats, but I was rebuked even when mentioning some of these same caveats. Caveats and exceptions were not being considered back when we were bashing the NSA daily...endlessly.

To state this position generally:  if we place cameras in places we might want to review later (on the police) and put in safeguards that the footage can be reviewed only with stringent restrictions that preserve civil rights, we may well end up with a good result when considering a more objectively truthful account of what happened at a given time. This good result can be contrasted with what we could come up with by putting together a version of the truth while we are effectively blindfolded and have liars in the room that know they can lie without being caught.

The simple logical outcome here is that more rights would be preserved than would be trampled upon if we have a filmed record of events we really really need. And significantly, in this case, police would not feel so free to mess around with citizens willynilly and they would know they would have a hard time lying. Or, we could put it even more generally: "The truth shall set you free."

This is not to say that our rights to strict privacy would not be trampled upon. They would. However, I have a lot of reservations about things that go on in private anyway. Beating me up in dark alley would probably be more effective than beating me on a well lit city street. Many of the things that have bit me in the past have come from private sources, from lies, or distortions in which I could have used the light of day. Privacy, to me, is overrated but that is just my personal experience. I need less privacy because I beat less people up.

There is, to me, some obviously pragmatic decision point that can determine when there is an overriding concern that trumps mere privacy preferences. If we film everything, then we can correctly make up our minds later so as not to deprive a citizen of their basic freedom from undue incarceration, for instance, and we could actually put in jail those people that we can accurately prove are guilty. To me, these trump my privacy when walking down the street eating an embarrassingly drippy ice cream cone. Oops, sorry madam, that was not me that dripped on your purse, I assure you it wasn't I. If only I had proof...

The deterrence factor for abuse by the police is quite a carrot for the ACLU. After all, they fight this sort of thing constantly. And absolutely, the police would be more likely to act professionally when their behavior could later be factually reviewed. The blue code of silence would effectively be overdubbed by a trustworthy narrator. Who cannot see that the pluses outweigh the minuses here? If the ACLU's methodology of handling videos is followed, the benefits would be phenomenal.

Bad things, like those in the following video, might happen less often. Yes, the video will be used in a lawsuit but what the woman surely wishes is that the police officer would have acted more professionally realizing the reality and ramifications of being on video and how acting unprofessionally  would drastically change his life. If he is a good man, he would be thinking more about his victim's future life. Don't even think the policemen did not realize the concrete slab was that hard. He knew she would be physically hurt.

---
Warning: this video is very shocking and violent. It contains violence, blood, and possible triggers.
Warning: this video is very shocking and violent. It contains violence, blood, and possible triggers.

---

When I argued my points about surveillance in general as being somewhat of a good thing, especially when used to fight terrorism or crime, I was roundly assailed in online comments. I know there are people who are no fan of the police or of a “police state.”  I can relate a few police horror stories of my own that happened when I was a young man.

So, here we have a chance to reform the police WITH surveillance. By the way, I used this very example of reform of the police in my arguments on the Internet to no avail. People were downright pissed with the NSA at that time and probably still are, while I was more generally happy to be alive to have the discussion due to the protection that might have been given me unbeknownst to me. The Bush financial collapse had made me realize my own precarious position in the world like 9/11 never did. Can we just put at least some bankers in jail? Nope, no surveillance.

Let me pick apart a few paragraphs in the position paper that seem especially telling to me. Again, ACLU, you have always been my main man!! High five, ACLU. I love ya and would proudly carry a card. No matter our slight differences at times, ACLU, you are da bomb. And despite my joking manner, I mean that most sincerely as a liberal who understands some things about the nature of evil accompanied by power. However, in reading about the history of politics and thinking things through, freedom as an abstract concept is not without its issues. Ask the parental victim of a child abduction. Would you have gone for a tad less freedom? More surveillance?

I am also struck by the "freedom" of buying guns at a gunshow in private, without the annoying background check. I want less freedom. I might even want cameras recording the event so we can id the people if they do something terrible. Do we have to hope that every criminal just happens to go in a Walmart to buy fertilizer or happens to go by an ATM when he is running from a robbery in order to id him?

The entire ACLU paper is reproduced at the end of this blog entry. Here is paragraph one of the excerpts I wish to analyze to get my mind wrapped around the issue of surveillance:
Although we generally take a dim view of the proliferation of surveillance cameras in American life, police on-body cameras are different because of their potential to serve as a check against the abuse of power by police officers. Historically, there was no documentary evidence of most encounters between police officers and the public, and due to the volatile nature of those encounters, this often resulted in radically divergent accounts of incidents. Cameras have the potential to be a win-win, helping protect the public against police misconduct, and at the same time helping protect police against false accusations of abuse. [-ACLU]
I agree wholeheartedly. We take a dim view and there are hairs to split about documentary evidence. But, I also could agree that there has never been much “documentary evidence” of rapes in prisons, for instance. There has not been much documentary evidence of pretty much any gruesome crime and no documentary evidence of the escape routes taken by criminals so that we might catch them and prevent them from committing other crimes. I think documentary evidence is great in court and in the investigative process. However, are we to make the statement that catching a policeman abusing her power, no matter how sickeningly disgusting that is, is much of a match to a mass murder, or terrorist act, and the successful prevention of those? Justification of surveillance cameras in general can be made in a great many ways. I am sure the ACLU has a lot of experience with police abuse of power… but I'm trying to point out that this is not the only game in town. A rape victim has far more experience with rape and the victims probably have this subject on their minds a great deal as well. A person who has a child abducted must stay awake in the night wishing for cameras everywhere along the path of the abductor.

Here is paragraph two:
We're against pervasive government surveillance, but when cameras primarily serve the function of allowing public monitoring of the government instead of the other way around, we generally regard that as a good thing. While we have opposed government video surveillance of public places, for example, we have supported the installation of video cameras on police car dashboards, in prisons, and during interrogations.
[-ACLU]
So prison rape is indeed considered. Before I hastily retract, I must say that this whole splitting hairs thing is a little on the frizzy side. We perhaps make an exception to the privacy of prisoners for what reason? Is surveilling a rapist here more important than in other places where rape is committed or along the escape route of the rapist? Is the ACLU taking advantage of the lack of rights prisoners have due to their confinement? I think not. Are the odds just greater at finding a crime here than elsewhere making it entirely a judgment call? I do see the point that monitoring the government is all that different from monitoring people, with the ACLUs methodology. If we are just being rhetorical, consider which direction the camera is pointed on body-mounted police cameras? At the policeman or at the people they are dealing with.  At the cashier or at the customer. And if the ACLU's restrictions of usage are followed to the letter, would it matter if I were filmed every day of my life if no one viewed it until it was to my or the public's advantage to have someone view it. I guess I understand the distinction the ACLU is making here about the police being in the spotlight, I just think that the line they have drawn is fairly chalky and artificial.

If I am lucky enough to have a surveillance video of a crime of which I am accused of committing, I can use this video in my defense against the government prosecutor gone mad trying to get ANY conviction (a fairly recognizable stereotype and abuse of power at least on television shows). Here, there is the same underlying need for protection against government overreach. Truth is important whether a policeman is present or not. I don't care where the truth comes from. Generally regarding public monitoring of the government as somehow more important than monitoring corporate greed when dicy mortgages are made or monitoring public transportation when I am coming home from work is not a clear case to me.

Here is paragraph three:
At the same time, body cameras have more of a potential to invade privacy than those deployments. Police officers enter people's homes and encounter bystanders, suspects, and victims in a wide variety of sometimes stressful and extreme situations.
[-ACLU]
Yes, CCTV cameras rarely enter homes except when the owner herself sets them up. Much of the ACLU paper addresses the safeguards that need to be in place. I really commend the ACLU for entering this area because, like it or not, sooner or later, we are going to need very strict rules of engagement here.
My point has always been that recording things that we may need later is not in itself a bad thing but an extraordinarily good thing if used in a good way. More criminals would be caught. More people would think twice before doing things that would turn them into criminals.

My intentionally extreme example would be to place cameras at all key intersections and parts of roads to record and digitally index the license plates of everyone going by. If someone commits a crime and we can surmise what car they are in… well, we are a lot more likely to catch them and they are a lot more likely to not commit the crime at all. If we only used traffic cameras for this single purpose then who could deny it would be a great leap forward in making society a better place.

However, if we allow local municipalities to use these cameras for routine traffic violations, to extract money from people for the local mayor to play with, then these devices suddenly become evil regardless of how many terrorists they might catch by chance. It is imperative that we make such cameras our friends, not our enemies by extremely strict rules of engagement that go out of their way to thwart abuse.

It is imperative that warrants for the viewing of data be linked to crimes that are obviously significant and against the public interest of everyone.  To that end I will not support the way the NSA has been run after Dick Cheney took them to the dark side.

However… we must be somewhat pragmatic in our decision making. That, I think, is what this ACLU paper is really saying except in a very specific example. It is painfully narrow in its scope but that is the point really. The less we actually need to view the fruits of surveillance the better, yet the more opportunities for later viewing of surveillance in hindsight when we actually need it, also, the better.




------------------------------------------------------------------------------------------------
Police Body-Mounted Cameras: With Right Policies in Place, a Win For All

October 9, 2013
By Jay Stanley
Download PDF version of the report (130KB)

1.Introduction
2.Control over recordings
3.Limiting the threat to privacy from cop cams
Introduction

When a New York judge found that the NYPD's stop and frisk tactics violated the constitutional rights of New Yorkers, one of the remedies she ordered was for the department to begin testing wearable police cameras, sparking debate and discussion of the technology there.

These "on-officer recording systems" (also called "body cams" or "cop cams") are small, pager-sized cameras that clip on to an officer's uniform or are worn as a headset, and record audio and video of the officer's interactions with the public. We have heard reports of police body cameras being deployed in numerous cities, and one prominent manufacturer told NBC that it had sold them to "hundreds of departments."

The ACLU has commented on police body cameras in the media several times over the years (and in stories surrounding the stop and frisk ruling), but the ACLU's views on this technology are a little more complicated than can be conveyed through quotes in a news story.

Although we generally take a dim view of the proliferation of surveillance cameras in American life, police on-body cameras are different because of their potential to serve as a check against the abuse of power by police officers. Historically, there was no documentary evidence of most encounters between police officers and the public, and due to the volatile nature of those encounters, this often resulted in radically divergent accounts of incidents. Cameras have the potential to be a win-win, helping protect the public against police misconduct, and at the same time helping protect police against false accusations of abuse.

We're against pervasive government surveillance, but when cameras primarily serve the function of allowing public monitoring of the government instead of the other way around, we generally regard that as a good thing. While we have opposed government video surveillance of public places, for example, we have supported the installation of video cameras on police car dashboards, in prisons, and during interrogations.
At the same time, body cameras have more of a potential to invade privacy than those deployments. Police officers enter people's homes and encounter bystanders, suspects, and victims in a wide variety of sometimes stressful and extreme situations.

For the ACLU, the challenge of on-officer cameras is the tension between their potential to invade privacy and their strong benefit in promoting police accountability. Overall, we think they can be a win-win—but only if they are deployed within a framework of strong policies to ensure they protect the public without becoming yet another system for routine surveillance of the public, and maintain public confidence in the integrity of those privacy protections. Without such a framework, their accountability benefits would not exceed their privacy risks.

On-officer cameras are a significant technology that implicates important, if sometimes conflicting, values. We will have to watch carefully to see how they are deployed and what their effects are over time, but in this paper we outline our current thinking about and recommendations for the technology. These recommendations are subject to change.

Control over recordings

Perhaps most importantly, policies and technology must be designed to ensure that police cannot "edit on the fly" — i.e., choose which encounters to record with limitless discretion. If police are free to turn the cameras on and off as they please, the cameras' role in providing a check and balance against police power will shrink and they will no longer become a net benefit.

The primary question is how that should be implemented.

Purely from an accountability perspective, the ideal policy for body-worn cameras would be for continuous recording throughout a police officer's shift, eliminating any possibility that an officer could evade the recording of abuses committed on duty. Of course, just as body cameras can invade the privacy of many innocent citizens, continuous deployment would similarly impinge on police officers when they are sitting in a station house or patrol car shooting the breeze — getting to know each other as humans, discussing precinct politics, etc. We have some sympathy for police on this; continuous recording might feel as stressful and oppressive in those situations as it would for any employee subject to constant recording by their supervisor.

True, police officers with their extraordinary powers are not regular employees, and in theory officers' privacy, like citizens', could be protected by appropriate policies (as outlined below) that ensure that 99% of video would be deleted in relatively short order without ever being reviewed. But on a psychological level, such assurances are rarely enough. There is also the danger that the technology would be misused by police supervisors against whistleblowers or union activists — for example, by scrutinizing video records to find minor violations to use against an officer.

If the cameras do not record continuously, that would place them under officer control, which would create the danger that they could be manipulated by some officers, undermining their core purpose of detecting police misconduct. This has sometimes been an issue with patrol car "dashcams" — for example, in the case of two Seattle men who filed a claim for excessive force and wrongful arrest. Parts of the arrest were captured by a dashcam, but parts that should have been captured were mysteriously missing. And with body cams, two Oakland police officers were disciplined after one of the officers' cameras was turned off during an incident.

The balance that needs to be struck is to ensure that officers can't manipulate the video record, while also ensuring that officers are not subjected to a relentless regime of surveillance without any opportunity for shelter from constant monitoring.

One possibility is that some form of effective automated trigger could be developed that would allow for minimization of recording while capturing any fraught encounters — based, for example, on detection of raised voices, types of movement, etc. When it comes to dashcams, the devices are often configured to record whenever a car's siren or lights are activated, which provides a rough and somewhat (though not entirely) non-discretionary measure of when a police officer is engaged in an encounter that is likely to be a problem. That policy is not applicable to body cams, however, since there is no equivalent to flashing lights. And it's not clear that any artificial intelligence system in the foreseeable future will be smart enough to reliably detect encounters that should be recorded. In any case, it is not an option with today's technology.

If a police department is to place its cameras under officer control, then it must put in place tightly effective means of limiting officers' ability to choose which encounters to record. That can only take the form of a department-wide policy that mandates that police turn on recording during every interaction with the public.
And this requirement must have some teeth associated with it — not only a risk of disciplinary action but also perhaps an exclusionary rule for any evidence obtained in an unrecorded encounter (for police who have been issued the cameras, unless there is an exigency to justify the failure to record). Another means of enforcement might be to stipulate that in any instance in which an officer wearing a camera is accused of misconduct, a failure to record that incident would create an evidentiary presumption against the officer.

Limiting the threat to privacy from cop cams

Most of the discussion around police recording has focused on its oversight potential. But that is only one of the significant interests implicated by recording. Equally important are the privacy interests and fair trial rights of individuals who are recorded. Ideally there would be a way to minimize data collection to only what was reasonably needed, but there's currently no technological way to do so.

Police body cameras mean that many instances of entirely innocent behavior (on the part of both officers and the public) will be recorded, with significant privacy implications. Perhaps most troubling is that some recordings will be made inside people's homes, whenever police enter — including in instances of consensual entry (e.g., responding to a burglary call, voluntarily participating in an investigation) and such things as domestic violence calls. In the case of dashcams, we have also seen video of particular incidents released for no important public reason, and instead serving only to embarrass individuals. Examples have included DUI stops of celebrities and ordinary individuals whose troubled and/or intoxicated behavior has been widely circulated and now immortalized online. The potential for such merely embarrassing and titillating releases of video is significantly increased by body cams.

Therefore it is vital that any deployment of these cameras be accompanied by good privacy policies so that the benefits of the technology are not outweighed by invasions of privacy. The core elements of such a policy follow.

Notice to citizens

Most privacy protections will have to come from restrictions on subsequent retention and use of the recordings. There are, however, a couple of things that can be done at the point of recording.

1.Recording should be limited to uniformed officers and marked vehicles, so people know what to expect. An exception should be made for SWAT raids and similar planned uses of force when they involve non-uniformed officers.

2.Officers should be required, wherever practicable, to notify people that they are being recorded (similar to existing law for dashcams in some states such as Washington). One possibility departments might consider is for officers to wear an easily visible pin or sticker saying "lapel camera in operation" or words to that effect.

3.Although if the preceding policies are properly followed it should not be possible, it is especially important that the cameras not be used to surreptitiously gather intelligence information based on First Amendment protected speech, associations, or religion.

Recording in the home

Because of the uniquely intrusive nature of police recordings made inside private homes, officers should be required to be especially sure to provide clear notice of a camera when entering a home, except in circumstances such as an emergency or a raid. Departments might also consider a policy under which officers ask residents whether they wish for a camera to be turned off before they enter a home in non-exigent circumstances. (Citizen requests for cameras to be turned off should themselves be recorded to document such requests.) Cameras should never be turned off in SWAT raids and similar police actions.

Retention

Data should be retained no longer than necessary for the purpose for which it was collected. For the vast majority of police encounters with the public, there is no reason to preserve video evidence, and those recordings therefore should be deleted relatively quickly.

Retention periods should be measured in weeks not years, and video should be deleted after that period unless a recording has been flagged. Once a recording has been flagged, it would then switch to a longer retention schedule (such as the three-year period currently in effect in Washington State).

These policies should be posted online on the department's website, so that people who have encounters with police know how long they have to file a complaint or request access to footage.
Flagging should occur automatically for any incident:

involving a use of force;
that leads to detention or arrest; or
where either a formal or informal complaint has been registered.

Any subject of a recording should be able to flag a recording, even if not filing a complaint or opening an investigation.

The police department (including internal investigations and supervisors) and third parties should also be able to flag an incident if they have some basis to believe police misconduct has occurred or have reasonable suspicion that the video contains evidence of a crime. We do not want the police or gadflies to be able to routinely flag all recordings in order to circumvent the retention limit.

If any useful evidence is obtained during an authorized use of a recording (see below), the recording would then be retained in the same manner as any other evidence gathered during an investigation.

Back-end systems to manage video data must be configured to retain the data, delete it after the retention period expires, prevent deletion by individual officers, and provide an unimpeachable audit trail to protect chain of custody, just as with any evidence.

Use of Recordings

The ACLU supports the use of cop cams for the purpose of police accountability and oversight. It's vital that this technology not become a backdoor for any kind of systematic surveillance or tracking of the public. Since the records will be made, police departments need to be subject to strong rules around how they are used. The use of recordings should be allowed only in internal and external investigations of misconduct, and where the police have reasonable suspicion that a recording contains evidence of a crime. Otherwise, there is no reason that stored footage should even be reviewed by a human being before its retention period ends and it is permanently deleted.

Subject Access

People recorded by cop cams should have access to, and the right to make copies of, those recordings, for however long the government maintains copies of them. That should also apply to disclosure to a third party if the subject consents, or to criminal defense lawyers seeking relevant evidence.

Public Disclosure

When should the public have access to cop cam videos held by the authorities? Public disclosure of government records can be a tricky issue pitting two important values against each other: the need for government oversight and openness, and privacy. Those values must be carefully balanced by policymakers. One way to do that is to attempt to minimize invasiveness when possible:

Public disclosure of any recording should be allowed with the consent of the subjects, as discussed above.
Redaction of video records should be used when feasible — blurring or blacking out of portions of video and/or distortion of audio to obscure the identity of subjects. If recordings are redacted, they should be discloseable.

Unredacted, unflagged recordings should not be publicly disclosed without consent of the subject. These are recordings where there is no indication of police misconduct or evidence of a crime, so the public oversight value is low. States may need to examine how such a policy interacts with their state open records laws.

Flagged recordings are those for which there is the highest likelihood of misconduct, and thus the ones where public oversight is most needed. Redaction of disclosed recordings is preferred, but when that is not feasible, unredacted flagged recordings should be publicly discloseable, because in such cases the need for oversight outweighs the privacy interests at stake.

Good technological control

It is important that close attention be paid to the systems that handle the video data generated by these cameras.

Systems should be architected to ensure that segments of video cannot be destroyed. A recent case in Maryland illustrates the problem: surveillance video of an incident in which officers were accused of beating a student disappeared (the incident was also filmed by a bystander). An officer or department that has engaged in abuse or other wrongdoing will have a strong incentive to destroy evidence of that wrongdoing, so technology systems should be designed to prevent any tampering with such video.

In addition, all access to video records should be automatically recorded with immutable audit logs.
Systems should ensure that data retention and destruction schedules are properly maintained.

It is also important for systems be architected to ensure that video is only accessed when permitted according to the policies we've described above, and that rogue copies cannot be made. Officers should not be able to, for example, pass around video of a drunk city council member, or video generated by an officer responding to a call in a topless bar, or video of a citizen providing information on a local street gang.
It is vital that public confidence in the integrity of body camera privacy protections be maintained. We don't want crime victims to be afraid to call for help because of fears that video of their officer interactions will become public or reach the wrong party. Confidence can only be created if good policies are put in place and backed up by good technology.

As the devices are adopted by police forces around the nation, studies should be done to measure their impact. Only very limited studies have been done so far. Are domestic violence victims hesitating to call the police for help by the prospect of having a camera-wearing police officer in their home, or are they otherwise affected? Are privacy abuses of the technology happening, and if so what kind and how often?

Although fitting police forces with cameras will generate an enormous amount of video footage and raises many tricky issues, if the recording, retention, access, use, and technology policies that we outline above are followed, very little of that footage will ever be viewed or retained, and at the same time those cameras will provide an important protection against police abuse. We will be monitoring the impact of cameras closely, and if good policies and practices do not become standard, or the technology has negative side effects we have failed to anticipate, we will have to reevaluate our position on police body cameras.

Published on American Civil Liberties Union (https://www.aclu.org)
Source URL: https://www.aclu.org/technology-and-liberty/police-body-mounted-cameras-right-policies-place-win-all