accomplish. we understand that it is not government that powers our nation but, rather, the ingenuity, hard work, and skills of the american people. your issues are our concerns. thank you, and god bless america. >> here on c-span, the communicators is next with indiana university associate professor mary gray discussing the ethics of personal data collection by internet companies. then, world bank president jim young kim speaks about climate change. at 8:00, today's justice for all march and rally in washington, d.c. >> c-span, created by america's cable companies 35 years ago and brought to you as a public service by your local cable or atellite provider. >> this week on "the

communicators" a look at personal data collection online and some of the ethics involved in that collection. joining us is indiana university professor mary gray, who was also a senior researcher with microsoft research. professor gray, are we being studied and researched when we go online? >> yes. simple answer, yes. in the sense that any company that is trying to figure out how to best serve their users, how to show them the best website, how to show them the best information from their friends, is constantly trying to assess what makes the best information service for you. at the same time, universities for the last 20 years, as we've moved more and more social has been moving to think of this as a social environment. this is the public square. there are plenty of universities and researchers like myself who have come up with, you know, over the years,

come up with research projects that have to do with studying people's experiences online. and in both those cases, there is as much interest in figuring out what it is that people do as there is in people figuring out how to navigate those spaces. so there's both wanting to know what people do and people wanting to figure out how best do i find information? there are two different threads f research that go on. >> how sophisticated is it? >> what part? >> the research part. >> we have basic tools for studying social interaction. there are fields of sociology, anthropology, to public health that have ways of studying how do people find each other, how do they find information, how do they share information? all of those fields have had to grapple with what do we do when a lot of that social activity is no longer happening in a neighborhood, in a park, in a nonprofit lobby or some social

space? so the tools we have are still -- they're still in development. we don't have easy ways to figure out how to study that social interaction and do it in ways that make it clear that we're there, make it clear that the information we're gathering is for research, so we're still learning how to do that. i think on the company side, companies have different investments for what kind of research, what kind of learning they're trying to do in looking at people's social interactions and they have a whole other set of challenges in doing that and doing it well and doing it respectfully. >> what are some of those challenges? >> i think the biggest challenge is that we are always moving between thinking of the information that's generated online as strictly a log, you know, a flat file of information like a search query or like a post and we treat that like it's just text.

at the same time what we are all grappling with is that it is not just a set of words. it's actually a conversation. so it's how to figure out how to study the information that's online as both information and communication. that goes the significant challenge. people have different expectations of what we do with say a letter, or a phone call or an in-person conversation. all three of those things are actively happening online in terms of people's point of view of what they're trying to achieve, whether they're sending an e-mail or they're texting, i.m.ing with somebody, instant messaging with someone, or having a skype conversation. all three of those are ways that we're generating information but they're also these moments of social interaction. >> is there a creepiness factor to all of this? >> the creepy question. it's a great question. as nk for all of us

someone who uses a computer every day we have certain expectations when we fire up our computers about who sees what we're doing, who we are sharing information with, and at any moment if the expectations i have are shifted because i realize there might be another party who sees what i'm doing, say for example if a message pops up and asks if i would like some help making a purchase, there are certain lines that we don't know we've crossed until it's too late. that's true for researchers. that's true for companies. there isn't a clear sense of what's creepy because that's so culturally specific. one person talking loudly on their cell phone in a park has no problem with somebody standing next to them on a bench and listening to that conversation. at the same time, you can have someone who is trying to have a private conversation and they will go to great lengths to be somewhere that's completely secluded. so we're not just dealing with

the cultural context. we're dealing with individuals' different preferences and experiences around privacy. and their needs for that privacy. >> let's bring joe marks of "politico" into our conversation. >> thank you, professor. so who is writing the rules of the road for what's appropriate online and is there a difference between the rules private rchers and for companies? with facebook, for example, most prominently they got into trouble recently by changing news feed results for certain people. >> right. the rules of the road for researchers are relatively clear, and i want to be -- i want to emphasize that it's relative. right now we have u.s. federal guidelines, often called the common rule, that set up the expectations for dealing with what's called human subjects research. in all cases of a researcher like myself doing anything that involves people whether dealing with data they've already produced like an archive, data base of a response to surveys, for example, the expectation is

that at some point if i was participating in that study i was given some opportunities to say, yes, i would like my information used for this purpose. so we have really relatively clear guidelines for how to deal with the offline world. the challenges we've always been playing catch up with what to do with the online world. these rules were written and put in place in the early 1970's. so no one was thinking about, well, how do we deal with a skype call? that said, researchers at every university who do this kind of research are constantly working on what are the best practices? that includes the ethical practices. how do you inform someone about the data that we might be looking at? i'll give you a really good example, which is it's really common to think about discussion forums say something that's like a google group, think of that as a public space because it's publicly visible. i don't have to sign up for an account. and at the same time, if somebody is participating in a google group, say just an open

group, they may feel fine with that being open with the assumption that maybe i am a fellow user of a particular group, say a tennis group. i'm not necessarily in that moment when i'm posting something to that group thinking, oh, somebody who is studying tennis is also reading this group and collecting information. so in that moment, am i actually interacting with somebody as a researcher when i read that discussion group and i collect all the discussion posts that are there? it's a genuine question. we don't really know. and so we constantly have to be thinking what might be the to find tions to ask out what is necessary in that space. one easy question, somebody who runs that online group posts something that says don't come here unless you ar member of this group. that is a real signal to a researcher they should seek permission. those kinds of clear messages are not always present.

as researchers have been working on how do i work with these different online environments, these social environments, we've had to continually bootstrap ways of seeking consent, of doing followup if we need to debrief someone who might be part of a study, and especially when talking about a globally istributed set of interlock -- interlocutors coming together it's even more complicated because it is not just the u.s.'s rules. for companies i argue there is already a set of rules in place and that is the federal trade mmission's comblines on good business practices. we have a very clear set of guidelines for what are the expectations a consumer has when they use a product. what we could be talking about is let the rules right now that govern human subjects research for researchers like me at universities really drive how we have this conversation about

what to do online and look at companies' involvement in doing testing, doing experimentation that has to do with product development, follow the guidelines in place for consumer rights. >> it seems the comblines in place for consumer rights, though, aren't necessarily jimbing with consumer expectations. how do you square that circle? >> i think the biggest challenge is for both those companies to be explicit and transparent about what it is that they're doing as they try and develop their products. in most cases any of the experimentation i know of that's happening in corporate settings is happening to improve a product. there are very few cases where someone is doing basic research in industry settings. my position at microsoft research is one of those exceptions. we're in a setting where we have access to resources internal to a company and that means i have to hold a higher bar for my research. in those cases if i'm

researching a topic that's both really interregular all to general -- integral to general knowledge i think i have a clear mandate and for me that is the common rule or the rules that regulate human subjects research. in the cases of companies that are doing any sort of experimentation, and the word "experiment" is a bit charged, doing anything that is studying user activity, i think they have to be both clear about their intention of who benefits from that activity if the consumer, individual consumer benefits from the activity by the improvement of their services, then i think there's room to be able to make sure those experiments are done with the greatest awareness and courtesy and respect to the consumer. i think there are plenty of cases where we also have the challenge to your point that most consumers do not realize how much information is gathered about them. that, to me, is the gap we have to bridge.

>> give some advice to a social media company that wanted to do an experiment like that to inform users appropriately, act appropriately in order to avoid the creepy factor. so often product development, the product in a lot of these cases is how you interact with your friends. >> exactly. >> which seems much more personal than buying a candy bar. >> that is a really great point. i think in many cases, in most of the cases, there would be an opportunity beyond terms of service. i don't think that's enough. there would be opportunities for companies to be able to say to a consumer before they've ever run an experiment, we'd really like to look at what you're doing for this reason. and be clear about the reason and give somebody the opportunity to say, i don't want to be part of that study. i think there are other cases where the question would not be answerable if someone is asked, would they like to participate. and usually in human subjects research we call that in cases where there's a compromise that we need to have some amount of deception to be able to ask the

research question. and deception in the sense of i need to be vague about why i am asking the questions about whether you like to play tennis, for example. in all of those cases it isn't as though a company doesn't know how to reach you. so if a company wants to do the kind of study that involves not seeking consent at the front end, there is no good reason they couldn't at the back end after an experiment has been run send a note out and say, hey. just so you know, we were looking at your data to get a better sense of your experience of this particular part of our service or our goals of improving our service, and if you'd like your data removed from this experiment, let us know. it would be very easy for them to remove that information. there are a lot of reasons given for why, what is called the debriefing process, why that wouldn't be possible. i don't buy it. i think there is plenty of opportunity here to do this in a way that respects consumers' autonomy and their right to be able to say, i don't want to

participate in something and be able to say i don't want to participate even if it's for my own improvement of service. >> can you talk a little about what the value is in this large, digital data set we have? what can sociologists and psychologists learn about the world that they couldn't 30 years ago because of this? >> this is so important because i think the hardest -- the hard conversation we need to have is that we absolutely need to be able to keep access to the interaction that's happening online precisely because so much of what we do as humans every day happens online. so the loss to the fields that try to understand human behavior, try to improve policies around support for online communities, a lot of my research, we need to be able to study people's experiences of the world as more and more of those move to a digital environment.

i think there would be such a tremendous loss if we don't work out these questions. >> can you tell us some stories about particular things that were learned? >> a lot of my early research was around young people identifying as lesbian, gay, bisexual or transgender or questioning their sexuality or gender how they use the internet particularly for young people in rural communities mostly in the southeast, kentucky, tennessee, west virginia. so being able to study, and what i learned from studying their experiences of going online, i learned what were the limits of what they could get online that really meant they needed to advocate for resources in their home communities. early 1990's to late 1990's the story was the internet will solve everything. we'll be able to just get a broadband connection, then it was a dialup connection, we'll have all the information we possibly need. that's not true. that has yet to be shown as the solution to social problems.

so in my research, and i think there are other examples of this, seeing the limits to what people can do with technologies and seeing what are the public resources that they need and private sector resources that they need to make up the difference, we can only learn if we're studying both parts of that puzzle, both the offline and the online. >> what were your tools? were you looking at google searches and things? were you sitting down with the kids? >> i was sitting down with the kids. so this was certainly before i had moved to microsoft research and most of what i had access to and i think there was another issue in terms of the data available, if it's a private company, i don't have access to the search logs. i don't have access to the back end. as a researcher, at a university, what i did have access to were finding young people who were using digital media. i did it two ways. i did a lot of -- i'm an anthropologist by training. i went to these communities.

i found those kids. they're not actually that hard to find. and i talked with them. i said, you know, who are your friends who are afraid to come to these meetings who are using digital media as their primary connection to the world? i talked with those young people and spent a year doing that research which is i think what it takes which is probably why i'm so afraid of situations where if we just stay at the surface level, the god's eye view of what people are doing online, we both risk giving people the sense that it's creepy and i want to stay away rom these worlds but we also risk not being able to access those young people or other communities through other means. >> professor gray, microsoft research is not a nonprofit place. >> no. >> how does your research benefit a company such as microsoft? >> i'm in a really exceptional and i think special

environment, because the expectation is that the research that i do, and right now i'm studying digital labor, so, for example, the project i'm doing on crowd sourcing right now, the goal is that i will be able to do basic research, answer basic questions, and that the contributions that i make to the scientific fields that can learn from what i gather will eventually be able to show better technologies, better routes to what companies like microsoft might have to offer. so there isn't an expectation that i am developing something specific. it's really a very old fashioned, basic belief that basic science advances all of us including technology companies. >> well, so many people that are watching this are thinking, oh, my goodness. i clicked on a website and all of a sudden tabs for that website started popping up. how pervasive is the big data? how much do private companies, microsoft, etcetera, know about

us? >> so, i think the -- being really precise here -- they don't necessarily know about you, the individual. what they have is a picture and the aggregate of what are a population of people, a specific age, a specific time, a specific location doing. what are they asking? so i think the challenge here is that, yes. i don't think anybody can deny this. there would always be a way to trace it back, trace that individual's search query to them individually. i don't know any company that actually collects information in ways that make it easy to identify a single individual in that aggregate picture. so when they're collecting data, they're really collecting it in bulk, you might say. it's very difficult to get to that information, so to be able to, as somebody internal to the company, i have to go through aappropriately a lot ofgatekeepers who protect the privacy of individual users.

none of these companies would exist if they were making it easy to access your private information. the challenge, and this is part of the conversation, is that most of the companies have only thought about the user's concern around privacy. so they've only taken up and are very good at safe gaurgding your social security number, your exact location, they won't divulge that information. they've worked very hard to keep others from getting that information. that's never bullet proof any more than somebody calling you at home and getting your social security number is not impossible either. in all cases, tech companies have put a lot of time into thinking about protecting your privacy, and that means that they don't know a lot about you as an individual that can be easily dished up. what they haven't had to think about is that you actually might care about your information being respected. but you might not want to see an advertisement that's related

to a recent instant message exchange that you just had. that's what they're really trying to figure out how is what's a respectful way of dealing with information, not just a way of protecting your information. and i think that there's a lot of conversation about what that could look like, but, yes. they know a lot about what all folks using their systems are doing, because they need to know that to improve their system. that, i think, in terms of educating the general public, making it very clear to the general public that any effort to give them advertising that's tailored to them, any effort to make it easier for them to see tweets from one set of people instead of another set of people, all of that energy means i have to collect a lot of information about what you're doing. >> what's your sense of what's driving that online privacy and online respect for information conversation?

because a lot of studies tend to show that people are very concerned about privacy if you want them to pay for g mail don't give up the privacy. >> it's fascinating to me. the pew research on internet and life just came out with a study not too long ago precisely about this. what are people willing to give up in terms of information online? most people's sense of what it is they're protecting when they online very much works from that privacy framework. we all tend to think about what do i worry about sharing as an individual. we haven't thought that much about, what am i uncomfortable sharing about our exchanges to people. when i'm going online and having a text exchange with, say, three other people, i'm not necessarily thinking about privacy in that moment. i'm with a group of people. in my mind, you know, i don't know if this is your experience, when i go online and i'm talking with a group of

people i'm not thinking about my individual privacy. i'm really thinking about our shared relationship in relation to the privacy or confidentiality of what we might be talking about. i might care deeply about keeping a conversation confidential like for the young people that i worked with. so i think the problem and the challenge is that right now we're thinking about privacy and kind of old fashioned ways that have to do with, say, a social security number, a home address, a location. and those are not necessarily the kinds of information that i might feel protective about when i'm thinking about the recent facebook study, when i'm thinking about what might feel intrusive about what a company is doing. it's a moving target for both the people using these services in terms of thinking what's the privacy that matters most, and it's a moving target for companies. >> very quickly, here are some pew -- some of the pew top numbers. 80% in the study agree they're concerned with government tracking.

64% think more should be done to regulate advertisers. 61% disagree that online is more efficient because of personal data sharing. 55% are willing to share for free online services. 81% don't feel secure using social media. >> oh, wow. i didn't remember it was as high as 81%. i think the -- i want to circle back to something you asked about before. the education around what it is that's shared online is key here because i think in most cases when people are thinking about their safety or their security online, they're becoming increasingly aware that the information is being collected and is not in all cases being collected to improve their experience of a service. it's to sell them a product. so i would argue we need to get to that place where we're not talking about a free service. it is not free. in exchange for a service, we are paying with the information

that we put online. so with any of these services, it really comes down to being able to say, is it worth it to me to be able to to have this account if i know that advertising in the same way that, you know, a telemarketer could call me at home, that i'm giving up information that will be made available to those advertisers? that's the exchange. it's not free. >> that is an interesting point. i was curious when you talked about the monitoring the amount of information being gathered is focused largely on product improvement. there are cases where in industries most prominently apple and google both offering encryption for smart phones, where privacy has become the product improvement. it's what people are selling on. that hasn't happened in most things on the internet. why is that? >> because i think a loft the money that's made for internet-based services comes from advertising. look, for most of these companies, and i think

crosoft is lucky it's in the advantaged position of not needing to rely on selling your information to be able to keep its business going. other companies that are primarily selling your information to keep themselves alive, any company based on an advertising revenue model, they need you to keep giving information and feeding it because the only thing they can really profit on is selling your, not just your personal information, but your activity. your social information. selling that to an advertiser. so i think there are companies that if they have other products and they can in many ways sell privacy as a luxury item, which i find is something we really need to discuss as a society, then they're going to have the advantage, because they don't have to rely on advertising dollars. in cases where a company has to rely on those advertising dollars it doesn't have the luxury of being able to say i

won't sell your information because that's what they have to sell. >> how is this going to change in the next 10 or 15 years as the internet as things developed? kenmore can be selling information about all the junk we take out of our 'fridge at midnight. >> yes. i absolutely can picture that day. i think that's why we have two roads. the future will look like an accelerated version of the selling of our personal and our social information or it will be a path where we make decisions as consumers about what we want sold and what we don't. the internet presents a new set of opportunities to have that conversation because you're right. as soon as you have -- as soon as you have the capacity to collect information on any activity that you're doing, there is going to be a desire to get that information. so it's going to be up to i think the public to be able to say these are the expectations i hold ransparency that

for companies trafficking or selling my information. unfortunately, until there is enough of a public expectation of both companies and researchers respecting their information, it's going to be a land grab. >> and, finally, professor gray, joe asked about advice for companies. what about advice for consumers? what would you advise to consumers who feel a little creep factor here? >> it's really tough right now. i don't think consumers have very many options. they absolutely should and always will have options with researchers because they can always come back to us and say that's not okay with me. in all cases, researchers need to be held accountable for their collecting of any information that involves human interaction. that's very -- that is made very clear. we have a mandate to do that. but in the cases of dealing with a company that might be collecting in fings and a consumer is trying to figure out exactly what kind of

information, they're not given very many tools to figure that out. so the most i can imagine a consumer being able to do is really educating themselves on how much of their interactions are tracked to benefit a product and they need to decide are they willing to have that tracking part of their experience online and really get out of the framework that it's free because they are the product in that moment. it's not free. >> professor, microsoft researcher, mary gray has been our guest. joe marks has been our guest reporter from "politico." >> c-span, created by america's cable companies 35 years ago and brought to you as a public service by your local cable or atellite provider. on this weekend's "newsmakers" john yu is our guest. he served as deputy assistant attorney general in the george

w. bush administration when he wrote several memos that are viewed as providing legal rationale for using enhanced interrogation techniques in questioning terrorism suspects. he talks about the recently watch the interview sunday at 10:00 a.m. and 6:00 p.m. eastern on c-span. world big president jim young cam about climate change priorities ahead of the lima, peru international conference. came at an event hosted by the council on foreign relations. this is about hour.