Friday, February 08, 2013

Podcast: Data and the law with Deborah Salons

Deborah J. Salons is a Washington DC-based attorney whose practice focuses on telecommunications, cloud, information security and privacy law.  She is also a Certified Information Privacy Professional through the International Association of Privacy Professionals.   With almost a decade of experience, Deborah is a skilled government affairs professional with a proven track record of leadership and coalition building in the technology, telecommunications, and media industries.  Deborah earned her undergraduate degree in Communications and Speech from the University of Washington, Master of Arts in Communications Management from the Annenberg School for Communications at the University of Southern California, and law degree from Indiana University School of Law- Bloomington.  While in law school, Deborah served as Editor-In-Chief of the Federal Communications Law Journal and clerked at the FCC and NAB.  Deborah is licensed to practice law in California and the District of Columbia.

I got to know Deborah when she was moderating a panel I appeared on at the Open Data Center Alliance's Forecast 2012 event. It's taken me an embarrassingly long time to get this podcast scheduled and recorded but I hope you'll agree it was worth the wait. This is an important topic but one that is, unfortunately, often poorly served by shallow or sensationalistic stories, blog posts, and papers. Deborah, on the other hand, takes us systematically through the legal issues associated with data at each point in its collection or use and offers lots of great, practical advice.

In this podcast, she talks about legal issues related to data, "big" and otherwise, and how concerns can be managed with the proper education and the proper teamwork between your legal counsel and your technology folks.

Listen to MP3 (0:28:10)
Listen to OGG (0:28:10)


Gordon Haff:  You're listening to the Cloudy Chat Podcast with Gordon Haff. Hi, everyone. This is Gordon Haff, cloud evangelist with Red Hat, and I'm here on the phone with Deborah Salons, an attorney who is focusing on cloud information security and privacy. I know that's a topic that's at the top of a lot of people's minds, so I think this is going to be a really interesting podcast.
Welcome, Deborah.
Deborah Salons:  Thank you very much, Gordon, for having me. I'm very excited to be here.
Gordon:  Great. Can you maybe introduce yourself a little bit?
Deborah:  Sure. My name's Deborah Salons. I'm an attorney based in Washington, DC. I've been practicing for about 10 years. I started practicing in the telecommunications regulatory space. As we all know, as technology converges and telecommunications converges with technology, I've been trying to follow that current, and my practice has been melding more into investigating issues in regards to cloud computing and information security as well as privacy issues.
Gordon:  This is obviously a topic that makes for a lot of, perhaps, sensationalistic headlines and fear stories and what have you. I think we'll maybe start off this discussion at a high level. Our legal concerns around cloud computing generally, and specifically around the handling of data, whether on‑premise or in a public cloud, is it just a bunch of click‑baiting headlines, or is there really something there that people need to be concerned about?
Deborah:  I think that there really is something there for people to be concerned about, but I think that those concerns can be managed with the proper education and the proper teamwork between your legal counsel and your technology folks.
Before I get any further, I do want to make the disclaimer that anything that I discuss on here is just legal information. It's not legal advice. I have to give the disclaimer since I'm an attorney. I'm not your attorney 
But I do believe that it's very important for technologists to collaborate with their attorneys on these issues, because they're not just headlines. Regardless of whether your data is stored on‑premises or in the cloud, depending on what kind of data you have and where it's located, you will be subject to privacy and data‑security regulations.
With those things can come civil penalties, can come enforcement actions, private lawsuits, criminal penalties in some cases, which is very scary, and FCRA also have some criminal penalties attached to that. The bottom line of all of this is it could be a PR nightmare for your company.
Regardless of whether the data is on‑premises or in the cloud, you're facing some legal issues. Now, once it goes into the cloud, there's another layer on top of that. Jurisdiction issues come into play, some data sovereignty. Of course, there's the security concerns that come with cloud, but you also have security concerns when it's on‑premises.
These issues are something to be concerned about, but like I said, if you start with your legal counsel at the very beginning, at the get‑go, I think you can really manage those concerns.
Gordon:  I think that's really a good segue into the next part of this discussion, because a lot of the debates or discussions around legal issues in the cloud can be very high‑level, and I'm not sure they're always that practical. But you've given a number of presentations that I've seen where you really go through specific actions that people can take in order to minimize their legal risk, so I'd like to walk down that path. The first step is, really, collecting data in the first place. We're really focusing on data here because, although those aren't the only issues, I think they're probably where the most severe legal issues can potentially arise. What are the issues around just collecting the data in the first place?
Deborah:  Sure. The data collection, it's probably one of the most important steps in the whole Big Data process, so to speak, because that's when you're deciding what information you're going to take in, and depending on what kind of information you have is what kind of regulations that you're going to be subject to. The problems with laws in this area is that you can't really take a book off of a shelf in regards to Big Data law. These laws come from state laws, federal laws, international laws, so it's really a cornucopia of laws that you're going to be subject to.
At the data‑collection stage, some of the concerns that you're going to have is, at the time that you actually collect the data, are people opting in? Are people opting out? Are you getting this data from your customers? Are they visiting a website? You have to be also very concerned about your privacy policy.
If you're collecting information on a website, and you say one thing and you're doing another, you're going to be in violation of some FTC rules. The privacy policy is something that you really need to look at through all the steps, but at the initial step, you really need to take into consideration what you're telling people and how you're getting the information.
The next thing I would say is you really need to also look at the kind of data that you're collecting. Are you collecting personally identifiable information? Are you collecting health information that's going to be subject to HIPAA? Is there financial information that you're collecting? In the case of telecommunications companies, is there CPNI that you're collecting?
Human resources, student citizenship, classified information, all of these things, whatever kind of information that you're collecting, at the point of collection, you should really look at what you're going to be subject to.
Finally, who are you collecting the data from? If you're collecting the data from children, you're going to be subject to the COPPA regulations, Child Online Protection Act. If there's state‑specific resident laws, if you're collecting information from someone in California versus New York, there might be different regulations that you're going to be subject to.
Furthermore, if you're collecting information from international citizens, there might be some other regulations that are thrown in the mix.
At the very beginning stage, at data collection, I would say your initial privacy considerations would be how are you collecting information, what kind of information are you collecting, and who are you collecting the information from.
Gordon:  One of the interesting things that point to it is in "big data" discussions. The default that a lot of people seem to think is good. "Hey, we just collect everything, and then all these wonderful insights are going to fall out that." Well, first of all, we've seen, often, those great insights don't necessarily just fall out. But I think what you're telling us here is that some types of data collection have definite costs or even potential risks associated with them.
Deborah:  Right. With all the capabilities that we have now, it's easier to collect more information. It sounds like it's probably better to collect more information than less, if you have the capabilities. You could learn more about your customers, and you could actually take some of that information and provide better services to them. But on the flip side, even if you're collecting it and it's not identifiable, there are ways, when you collect so much information, that you could perhaps re‑identify that information. That might trigger some laws, or that might go again their privacy policy.
Again, that's very, very important, because if you are doing something contrary to what you're telling people you're doing, and it could be even on accident or it could be that you didn't intend to do that but you're able to do that, you could run into a problem.
It is a very interesting balance between the excitement of having all those capability and technology to collect this information, run analytics, and get all this information about your customers or what have you—but  on the flip side, you might be running into some problem. Again, this is another situation where you want to have the technologists, the lawyers, and the business people all come together and really figure out a plan and see what the best road map is avoiding potential risks.
Gordon:  Deborah, as we move to the next stage, where you have all this data. You're now adjusting it and cleaning it. You may be getting multiple data sets together. This is certainly what the areas where this whole issue of "Is the data personally identifiable or not?" really comes in. In fact, you see more and more news items or research papers and the like talking about how some types of data that people have just been assuming aren't personally identifiable increasingly turns out they really are, because it wasn't that they couldn't be personally identifiable. It's just people figured it was going to be so hard that they were safe.
But it served like "security through obscurity." It doesn't necessarily work that way.
Deborah:  Right. I think that this is becoming more and more of an issue, because even if you collected a whole bunch of data that was not personally identifiable, the more data you collect, the more dots you can connect. Then I think that that is the downside to all of this. You might be collecting a whole bunch of information that you might not think you would be able to match up with a certain individual, but once you collect the whole bunch of data points, you might be able to reverse‑engineer, so to speak, the identification of that information.
That's something definitely to be on the lookout for, because with all of this technology, the capabilities of the Cloud, and computing power, it's a reality.
Gordon:  We've got this data now. We've brought it into our systems. We've hopefully cleaned it up somewhat. Now, we're wanting to actually use this thing, whether it's people looking at it, data‑mining type of techniques, or typically some combination thereof. We're now wanting to get something useful out of it. What legal risks can come in at that stage?
Deborah:  Well, I think, when you have the human element of just who's handling the data. It could be a computer, it could be a human, but at the end of the day, you have humans giving computers commands. But some interesting aspects to this is that I found, and I'm not an import/export specialist whatsoever, but speaking to some of my import/export attorney friends, one of the things that they were discussing with me is citizenship and work visa issues with some of your employees working with certain kinds of information.
It could be technical specs. It could be classified information or business trade secret type things. But you could actually, with the combination of the type of information and the citizenship or work visa classification that your employee has, certain combinations can trigger import/export laws, and you can actually make an illegal export of information with your information being on American soil and your employee being on American soil.
Again, I'm not a specialist in this area, so I'm not quite sure what that combination is, but I thought that was very interesting. That's something that all companies should be aware of because you don't want to be in a situation where you had that combination and you're triggering import/export laws when that might be something that's not even in the purview of what you're looking at.
You're thinking, "Oh, I've got data. It's all here." You would never think of import/export which I think that's a very interesting twist to all of this.
On top of that, just having people having access to certain data, regardless of their citizenship or where it's at, you really need to look at your corporate policy to see who has access to certain data and what their limitations are. Anyone in the security community will tell you that a large number of data breaches occur internally and occur within employees within a company.
It might not necessarily be a hacker but it's someone internal. So you really need to see what your constraints are, what your corporate policies are to ensure that the people who need to see the data, see the data and the people who might not need to see it don't have the access to it. Not only do you need to look at information security from an outside hacker situation, but you also need to do your due diligence and make sure that you're handling your information security internally as well.
Gordon:  We've done something with this data. The storage is cheap these days, so I think in general people you're serving claim to hold on to this data they've gone to a whole bunch of trouble to collect and analyze because who knows what could be useful a year from now. But that introduces some risk as well.
Deborah:  Sure, and I think storage can be cheap these days, like you said, and it could be the tendency of some business people to look at cost and go perhaps with the cheapest storage or whatever is the most convenient. But you really need to take a step back and look at the big picture in regarding storage because, first of all, you need to figure out where you're physically storing the data because even though the cloud seems like it's borderless, it really isn't. There's jurisdictions and either the location of the data really do make a difference as far as what regulations you're subject to. That could be a huge business consideration. And so, there's national security considerations. There's the EU data privacy laws and data sovereignty issues.
I know the European Union, they view anything that has to do with the U.S. cloud computing provider is subject to U.S. Patriot  Act and that's something that they're not comfortable with. So, as far as where you're physically storing the data, that can make a huge difference. Then during the storage, you need to understand how you're protecting it, who has the duty of care.
If you have an outside storage provider, what kinds of firewalls? Encryption standards? What are they doing to protect the data, as well as what you're doing?
That comes to the question of what if there is a security breach? That could be a huge problem as well, because breach notification laws are different in almost every state in the United States. You might have to, depending on where the information is located and who you're collecting the information from, you might need to make specific notifications about breaches to different people.
All of those notifications might have different caveats and different specifications, as to how you need to notify people. That could be 49 different notices [corrected from 48 in the audio] that are going to be going out. On top of that, another interesting thing that's going on with breaches is that the SEC, Security and Exchange Commission, is now interested in breaches.
If there is a large enough breach in your system and you're a publicly traded company, you might need to disclose that information to the Security Exchange Commission. Because, in the end, this can affect your business. So that's a new string of laws that's also developing. Breaches are very important, not only to your bottom line, but to your PR.
Then also on a legal stand point. Where you store the information and how you're storing it is very important. Don't just look at the cheapest avenue. You have to really look at the big picture.
Finally, I would say, how long are you keeping this data? You really need to balance out. Like you said, you can gather all this information.
You can keep it forever if you wanted to. You really need to figure out how long you really need to keep the data and what the risks are, attached to keeping the data for X amount of time. And, when you dispose of that data, obviously, you have to be responsible about that. Do that in a responsible manner, so you're not breaking rules at that point either.
Gordon:  What are the implications for things like e‑discovery, with having these much larger data sets?
Deborah:  From what I understand, and I'm not as much of an e‑discovery expert, the more data you have, it might be harder to sift through things to be responsive to discovery requests. I was reading an article earlier today that said that could just be the tip of the iceberg.  When you actually have these discovery requests and you have all this data to be going through, there can be things that pop up about how you're storing your data, what kind of corporate policies you have. In case you have customers, if you're following your privacy policy or doing what you said you were going to be doing.
So on top of just the e‑discovery issues of being responsive to discovery requests, things could come out in the process that would not be favorable to your company. Not only is the e‑discovery an issue. But also having your corporate policies in line before you're in a reactive situation with e‑discovery is very important because some of these things can creep out in the process and those processes become public, with being involved in litigation.
Gordon:  Finally, in the case of researchers, companies under some circumstances, at least some of their data, they want to publicly share in some way, shape or form. Obviously, they know about medical records, confidentiality and so forth, but what are maybe some of the less obvious things that companies need to think about when they make data public in some way?
Deborah:  I think, when you're at the point when you're ready to share and publicize the data that you've collected, you're at the end of the whole Big Data process. I really think you need to wrap back around to the very beginning and look at the privacy policy and the agreements that you made with the customers and the people that you've received the information from. Because, at the end of the process, people might not look back at what they said that they were going to be collecting and how they said that they were going to be using it.
I would advise people to make sure however you're sharing and publicizing this data is, in fact, in tune with what you said in the very beginning, because you can get into a lot of trouble with unfair and deceptive trade practices, with not matching up with your privacy policy or what you said you were going to be using the information for. You have to really make sure that the end matches up with the beginning.
Obviously, with the types of information, you don't want to be spilling health information or classified information or things like that. But you really need to see, can you really legally share the information? Is what you're doing not only consistent with your privacy policy, but is it also consistent with the rules in your jurisdiction?
Also, if you're sharing unidentified data, you really need to look at, if I were to share this data set and it's unidentified, is there a way that it can be re‑identified? Because you don't want to be liable for that at the end of the day if there's a problem doing that.
This whole big data process is a great thing and you can find great information that would be very useful to the public. But you just need to make sure that it's in line with your privacy policy, that you're doing things on the up and up. And see what could happen. After you share that, what could be the next steps. Because you need to look into the future, to make sure that you're not going to walk into any kind of a problem that you just didn't think about looking in the future for problems.
Versus what you did in the past with your privacy policy and what you're doing at the present time, with the legal regulations. And, what you can disclose.
Gordon:  Even with ostensibly public data, we've seen issues recently. For instance, registered gun owners' locations being published. Even though that was, ostensibly, public data a lot of people weren't very happy about that. It doesn't seem, in many cases, that we've really come to grips with the implications of certain types of public data. They're no longer gets filed away in a dusty town hall somewhere. They can be accessed with the click of a mouse button and be mashed up with other types of public data, to really expose a lot of facts about people that they don't necessarily feel comfortable being something that their neighbor can just go online and look up.
Deborah:  Sure. It's really about connecting the dots. That's a big issue. I think that that's something that the courts are actually starting to recognize. I know that I've been in social engineering presentations where you can connect dots from people's social media websites and other things on the Internet that's free, and it's public information. Once you have a whole bunch of different dots you can connect them and figure out things out very, very easily.
I know that the Supreme Court was just looking at this issue in the Jones case, which was a case that happened here in DC, and it was in regards to legal searches, searches and seizures and whatnot, and if a warrant was needed for sticking a GPS sticker onto a car to see where that car went.
The interesting thing that happened in the Supreme Court and also in the DC Circuit Court dicta was that the justices were discussing about having an officer doing surveillance on someone. You can see where they go from one point to another, but it might not be 24 hours a day, seven days a week.You might see someone go into a grocery store. You might see someone go into a doctor's office.
But if you have a constant GPS sticker and you see someone went to the store at this time, and then someone went to a doctor's office at this time, like if it was a gynecologist's office, for instance.--this was actually discussed in one of the court cases--and then someone went to a children's clothing store or a baby crib store, something like that. When you put all these things together, you can figure out that the person is expecting a child, and it might not have anything to do with the case.
It's almost like the more information you have and the more dots that you can put together, it provides you with a bigger picture than if someone was just following someone in a car and you see someone go one place, you see someone go another place.
You might not have the time. You might not know who's there. You might not know a whole bunch of other different things.  But the more data you have, the more dots you can connect, and that's when things get tricky.
Gordon:  I guess to maybe wrap up here, we've talked about a lot of different regulations and you've talked about some of the court cases. How well established is the law in these areas overall, and how consistent is it between different jurisdictions, both states and internationally?
Deborah:  Sure. I would say that the law certainly has not caught up with the technology, and I think that that's very common when you're dealing with new technologies, whether it be in the wireless arena or computers. The law is always slower than the rate of innovation. I think the problem is, you have all of these new technologies and the laws don't really match up with what you're dealing with. That is a huge problem. That's why I really believe that business people, the technologists and the lawyers all really need to get together. Because you don't want to find yourself in a position, down the road, where you're facing legal problems because you didn't think about it. But at the same time the regulators weren't thinking about it either. You might come down the road where something affects you that you didn't think.
Also, if you're a big company with a big idea and some kind of new technology, it's also important for you to be talking to your regulators and rule makers and lobbying Congress and whatnot. So that you have a say and you can provide information and let people know what you're doing and how the laws can be changed or developed, in order to help your business thrive.
But as far as the rules being uniform, that's also a problem. Like I mentioned with the data breach laws, they're all different. There are 49 different data breach notification rules on the books. [Transcript note: 46 states, DC, Virgin Islands, and Puerto Rico (making a total of 49). Corrected from "48" in the audio.] So that is a problem. Because if you have a breach and if you have information from people in different states and you have information located in different states, you might have to send out 49 different letters.
That can cost a lot of money, with legal fees, to be frank. I know that that's something that the government is looking into making consistent, here in the United States. That is a problem.
Another problem is, with different countries, obviously, you're going to have different rules and regulations. Just as a philosophical matter, the European Union and the Europeans look at privacy a lot different than we do here in America. The Europeans look at privacy as a human right, and we look at it as a consumer protection law.
There's a different perspective at it and how people value privacy is different, so their laws are going to reflect their values a lot differently.
In the EU, I know right now the data protection authority is making moves to try to have some kind of consistent law with all the European Union nations in regards to privacy. Right now, they can all have different laws, just following under the directive, but they're really trying to get the regulations to be consistent so that at least within the European Union there can be some consistency there.
They're currently working on this, but it hasn't happened yet. I think between countries and even within the United States with the states, there is a problem with consistency. That's all the more reason to have your legal counsel involved, because you might think you know the law in one place, but it can certainly be different in another.
You want to make sure that the way that you are crafting your business and your business model and how you're doing things, it's the most legal and you are avoiding any kind of expensive legal fees at the end of the day, whether it be dealing with all of these different regulations, or dealing with litigation, which is obviously the thing that you don't want to have happen.
Gordon:  That sounds like good advice. Thank you very much, Deborah.
Deborah:  Thank you very much, Gordon.

1 comment:

Unknown said...

Informative.Thanks for sharing