data and the raw data. he said based on the work that he's done it is not [ indiscernible ] -- anyone who had access could have had access to all the data that they work with. and i know from the email exchange that we publish, the documents [ indiscernible ] work as well. there's a next step to be taken. during the campaign and that's that [ indiscernible ] all the information made available to them using the platform. so currently all available data was ready to be sent to breitbart. [ indiscernible ] they needed to do. you know from facebook that they found evidence of shared audience files on sel and i.t. campaigns that they had run and the correlation [ indiscernible ] and aiq said they had been given all the data by sel. [ indiscernible ] if they are, i am not sure the data was truly reflective of the nature of the relationship between aiq and [ indiscernible ]. >> i think as i explained to mr. chairman, was that the example that i gave you was that it's not -- it would not be unusual for kombucha ana

now-defunct british data firm. he was joined by a political science professor from tufts university and the director of telecommunications studies at the university of florida. this is two hours and 15 minutes. >>> thank you to everyone for coming. the facebook matter ining alexander cogan and cambridge analytic, shed a bright light on the data practices of some of our largest technology companies. although advertisers and political campaigns have collected and used to data for years, the public seemed generally unaware. this story has forced both the public and lawmakers to confront serious issues that need to be addressed, incding the role congress should play in promoting trans-parity -- transparency regarding collection and use while ensuring a well-functioning marketplace for our data dependent technology to drive further innovation. we started thatconversation with mr. zuckerberg last month. i hope that today we can continue a productive and meaningful debate about these serious allegations. unfortunately, ev

data that has been collect it. do you think that would behi helpful? >> particular harms are two critical pieces for any privacy directive. i wonder if overly focusing on data portability is an issue. there is disbelief that the able to take platform and get to platform be is beneficial and increases the likelihood of data breach of the platform. in addition to data portability with a strong requirement of insecurity and data safeguards on each platform. >> thank you. mr. battelle come you brought up the issue of elections. is that right? so, the honest ads act is other billing doing with senator mccain and senator warner and not the book you what happened in the last election where there was no disclaimer requirement in no disclosure requirements and basically say you have to have the same rules with the tv ad for tv, print ad for print. some of the companies for some reason are doing it on c their own, which we truly appreciate. again, we will have a patchwork of rules if we allow them to do that because they are

scrape data from social media platforms and they're misusing the data it's being alleged that this data was used to. raid an election in america and twenty sixteen so how do you deal with the fact that at the moment data got a p r problem that you know if i can put it in those terms yeah you can and i'd like to make that more specific centralized data has a p.r. problem right it's something that's out of control the end user people are giving away their data for free to facebook and other big companies google who are kind of taking advantage for that but of course that's their business model that is actually the current and only business model of free services on the internet could that be changed could we put the person the source of the data back in control so that they could decide who gets access to the data how it gets monetized maybe they earn a share of that all this is possible with technology but it hasn't been possible in previous generation products here is a question. you probably don't know the answer to this question but i'll ask the question anyway you know it's been esti

so this is data on, for instance, loyalty card data, transaction data.this is data that pertains to lifestyle choices, such as when car you drive or what magazines you drive -- i beg your pardon, what magazines you read. it could be data on consumer habits. and together with some demographic and geographic data and obviously voter data which is very important for the u.s. politics. [ inaudible conversations ] >> well, obviously the gsr data set. but i think you need to understand that that data was in the grand scheme of things was really a very, very small component of the larger data set. >> okay. >> thank you very much. just before we do, can i just clarify one things. because these newspapers aren't very recent. when you appeared in february, so much has gone on since. you made the distinction between cambridge analytica and the scl group as certain entities. it is the case that all of the companies have gone into either chapter 7 or administration? >> that's correct. >> so as far as the stl group companies. >> uh-huh. >> a concern, has it been the c

essentially data portability and data aggregation tool. facebook sudded anticipate claid copyright. when it threatens their service they go after it but when they have benefit from those partnerships they will allow access. >> thank you. thank you, mr. chairman. >> senator markey. >> thank you, mr. chairman. thank you for this excellent hearing. as former chief technologist at the federal trade commission do you believe that facebook violated its 2011 consent decree with the federal trade commission during this cambridge analytical debacle by sharing users friends information overriding users privacy preferences? >> so, i, of course, have to preface this by saying the opinions here are my own in my personal capacity not of the ftc or any previous employer. what i can say, the facts around the cambridge analytical data sharing was similar to the facts in 2011. ftc complaint alleged that facebook first shared private information like friendless to be made public when they were made private. told users they can restrict data sharing like friends only. when friends only didn't prevent i

data that has been collected. do you think that would be helpful? >> these are two critical pieces. conjunctival reliefs and are two critical pieces. i wonder if overly focusing on data portability is also an issue. there is this belief that being able to take your data from platform a to get to platform be is beneficial. from a security perspective, you are increasing the likelihood of a data breach for each platform. i think we need strong requirements for security and data safeguards on each platform. senator klobuchar: thank you. mr. patel, you brought up the issue of elections. the honest ads act is another dull am doing with senator mccain and senator warren. i would look at what is happening in the last election while there was no disclaimer requirements and there were no disclosure requirements. and basically so you have to have the same rules of the game that you have vincent audio ad for radio or tv or print ad for print. i don't think it's that hard and some of the countries now for some reason are doin

examining data privacy. earlier this year, reports surfaced a personality test application called this is your digital life founded in 2014 that was hosted on facebook and downloaded by 300,000 facebook users. who consent to the collection of data from the profile. however, facebook rules for third-party applications allow users to consent on behalf of their facebook friends for their profile information to be collected as well. this arrangement allowed this application to collect data from tens of millions of facebook users by getting a consent of only 300,000 users. the application also shared this information with a data analytics firm called cambridge analytica , which worked with political campaigns on targeting advertising. it is alleged that other applications participated similar collection practices but facebook changed the terms of service to prohibit this practice in 2015. what facebook sought from cambridge analytica and developers of personally tested applications that user data in question was

we thought collecting peoples data like we did was completely normal, accepted, and people whose dataas being collected new it was regularly happening. , my view isalcy perhaps one of the most concerning about the whole situation. it means that people and companies have done and will do projects where data will be moved and used in ways sooner appreciate. indeed, that is what we have in act. an unregulated industry with people working in ways that we now have to seriously wonder if the general public is ok with or even knows about. the industry does not have the same reputation as finance. culture of ask forgiveness and not permission is pervasive in tech. i see two core possibilities. people may feel angry and highlighted if they think their data may have been used as part of a mind control effort. cambridge analytic doused their internet to them into voting in a way they would not have. the data is entirely ineffective for micro-targeting. there is value to the data if you're trying to understand the general tends of great -- of the groups of people. for any one person it does not w

in facebook data or data from other sources doesn't help very much. persuasion affects decay rapidly. resear hasn at an ad may change a voters mind for a fleeting moment that the effect goes away amidst all of the other ads, news, posts and stimuli that accompany an election season. second, persuasion is hard because it is actually very hard even with top-notch data to figure out which voters are persuadable. the reason for this is, no person is persuadable all the time. persuade ability is not a stable disposition. you may be persuadable now but not tomorrow. persuasion is hard because when you try to figure out who is persuadable, your prescript -- predictions of who is persuadable is often important -- is often imperfect. campaigns often want to predict the racial identity of a voter. predictions of which voters are black or hispanic are wrong about 25% of the time. when a campaign sent the message targeted to these motors, one quarter them will be missed targeted. research suggests that voters penalize candidates who missed target them. they will

he denied this firm had ever received that data, denied his firm had ever used that kind of data for their work. today in a parliamentary hearing room in britain he admitted actually he'd been mistaken in his earlier testimony and actually, ah, thatas the data that they used. >> the chairman asked you, "they've not supplied you with data information?" your answer again was "no." and the chair then asked again, "your data set's not based with information you've received from them?" again you said "no." the chair, "at all?" and your final response was "at all." so since then we've had some rather conflicting evidence from dr. kogan himself, from christopher wylie about what data was supplied and used. indeed mr. wylie described that data obtained via the kogan act from facebook as the foundation data set of the companyhich may have collected data on up to al clearly the been ac by facebook following these revelations or allegations. do you want to clarify or amend now the evidence that you gave in february with respect to data supplied by gsr and dr. kogan? >> thank you. i'm grateful f

data? much like, you know, you look on facebook, and it says, you know, "the following apps can access your facebook," or access your twitter. the following people can access your data. i type in, i this old, i live at this place, and this zip code, i'm interested in that, and i can revoke that at any time, or i can go in and change it, or just say, you know what, i don't want anybody to know my age anymore. is that a thing we could do? christy: so, i'll break that into a couple of different pieces. some of those facilities exist within facebook, within twitter, if you haven't already done it, it would be fascinating-- scott: within twitter, but i'm talking about like a, you know, everybody in the world just accessing scott mcgrew's data point. christthere's a number of different companies that are trying to figure that out, not just because it protects your privacy, but because are making billions of dollars on monetizing your data, what should be some percentage of that going back to you?

moving forward, as the cdm program matures, we get more data and telemetry data, that's going to allows to do much more of a rich proactive set of hunting. one of the things we see across our mature commercial clients is not waiting to be attacked, not waiting for an event to pop before you do instant response, but employing hunt teams to go out and proactively looked for advanced adversaries on the network. i think the cdm program is a catalyst and enabler to that next level of, you know, capability across dot-gov. >> charles, anything to add on that? >> from a case load perspective, what we found most of the organizations we worked with learned about a breach because a third party told them about it. fast forward by several years and today what we're finding is about half of our clients are self-detecting breaches and the other half learn about it from a third party. so i think the better data that they have in order to try to practically hunt in their environment would help them more quickly be able to detect attacks in a much shorter period of time, and it's always better to be abl

accurate data with the ability to process that data. when it comes to science and technology those are investments to get more data and better data to process it that gives us a much more safe environment and handed that to the agencies that actually do ssa. >> the gentleman from colorado. >> thank you for putting together this hearing into the panelist for what you are doing for our country. general, the standards work for space situational awareness for space command how does this policy affect that work? >> the work that we need to characterize threats will continue through strap, in my commander who is also based in colorado. that work will continue it has to but what will change looking into the future it is a natural progression that we will have to partner closer not only with nasa department of commerce and transportation but the commercial sector as well there is a number of entities that do this mission and they have the capabilities and we need that leverage as well. it cannot be one-size-fits-all but take the best and integr

does that mean the data analysts will become the new star players?at this year's world cup finals in russia, data is set to play a bigger role than ever before. like it or not, data is already an integral part of modern football. you can't watch a televised game without being bombarded by all kinds of statistics. for the first time, fifa are providing an all—in—one technical package so technical staff can analyse data throughout matches in real—time. the snappily titled electronic tracking and performance system — or etps — will provide data on player metrics, such as distances run, speed, and position information. crucially, each of the 32 countries competing in the tournament will be provided with two tablets — one for data analyst in the stands, who will be able to communicate with the second device user on the bench, possibly a coach or assistant manager. we are using an optical tracking system. so we have optical tracking cameras in all of the stadiums. we know where the players and ball are at any time of the match. we have different metrics av

very large data centers.l right. ben cutler, microsoft research, thank you so much for joining us. fascinating. airbnb has frozen a major portion of its home listings at the behest of local regulators in japan. the company said it would compensate travelers who find themselves stranded. it warned hosts it would be list those that had not registered would delist those that had not registered for licenses as required. lawmakers with the cloud to regulate facebook after it revealed it had data sharing partnerships with chinese device makers. we will discuss next. this is bloomberg. ♪ emily: activist hedge fund janus partners has named facebook one of its top positions. it reflects a bigger theme in tech stocks, the best-performing group in the s&p 500 this year, despite a strange relationship with the trump administration and the growing threat of more regulation. joining us to discuss, caroline hyde in london. dan in new york. the latest controversy in new york -- in facebook, that they shared information with

could get data at the commerce department may have more data that dod may not have shared that data with either or both. we don't want to limit the idea only the government could do it we need to have partners that can share sometimes the challenges international data they may not want to give it to strategic command but will share with commerce for that is another reason would even civil authority rather than simply the department of defense a lot of countries around the world want to partner with united states air force but they want to partner with you because we are a separate faiththna agency capabe of science and technology apart from military capability. >> a lot of the private sector. more than happy to share and they already do in many cases. >> i hope they would share but i cannot guarantee that. >> i yield back. that unless we come to grips with space debris challenge, and now talking mainly about the catalog in the private sector to help us catalog the problem to define what that is. but have we given any thought to the private sector once it is catalog to do something a

everyone to have data sharing contracts. few examples that i would consider to be slightly over those interpretation where, a fund manager may in a very ed advice deliberate interpretation of the ules that says they are not sure they can actually provide evaluation data back to the behalf of the client unless the client has told the so. manager income do that points to perhaps a contract when this client first might have been three or four years ago being out of date. few u are finding a circumstances like that. the way to change that is to make sure cts and that you are not checking things ou don't need and you are very clear on what you are getting when and why. >> what is the cumulative impact so far of this? good thing byis a the way. i think all of this is a good thing. effect on people we work with financial advisors and fund managers and our job in is to put y business the two parties together. they have been doing business together a long time. helping them to communicate more closely. he key impact is effectively in

so just housing the data don't give anyone access to the data. we worry about the control. we worry about the use of the data for the purpose, and we monitor that through irb processes as well, so, you know, we've set up certainly an enterprise sensitive to the use and protection of the data for the very reason you remarked. >> you know, with you putting in all those parameters to protect the information, are there any challenges to accessing the complete medical records of veterans whe neen be? i guess i would want to know is it easily accessible, quickly accessible in situations where it needs to be? >> so there are two parts to your question. technically it is easy to access now. in terms of the tools, the the infrastructure that we've set up, hardware and software, the learning environment. what is still a bit of a challenge is the irb process. youkn, what we've been doing here is new. every step we take is new for everybody in terms of how we access data, and i think as we try and create the irb structure for accessing veterans data, we're sensitive to the fact that ma

analytics or data science. i put this in here specifically because in indiana we have been able to explore and do some things around data analytics and data science on the human population, health and human services area. and the architecture that is in place for those types of systems are tuned and they are use case specific for very large data sets that have to have algorithms run across them to produce these predictive analytics. that technology basically what we're trying to do when you were talking about security data and trying to get at least somewhat ahead of the things you want to get ahead of. i say here challenge your operations teams to not box your security teams in to just what we have available for the agencies use a a report for the other entities within the organization. instead look towards building architectures that mimic more of your data science, data analytics practices and let that be your storage mechanism and your capture so that you can get the increase flexibilities and the efficienc

and we need better data and accurate data and the ability to process that data. and so when it comes to science and technology, which is what nasa does, those are investments that we make. we want to be able to get more data, better data, and be able to process it in a way that ultimately gives us a much more safe environment and then take that and hand it to the agencies that actually do ssa and stm. >> thank you very much. thank you, mr. chairman. >> i think the chair now recognizes general from colorado. five minutes. >> thank you, mr. chairman, and thank you for all the chairman for putting together this great hearing and thank you to the panelist for what you're doing for our country. i have one question for each of you. general hyten, a lot of the data standards work is done at colorado springs at air force space command. so how will this policy protect that tremendously important work? >> so the work that we need in order to characterize threats will continue and will continue through stratcom, through my joint base commander who is also the air force base

it is up to the data analyst are they want to use this.ble for technology differing between countries, fifa. .. technology technology differing between countries, fifa... technology will bridge the day to divide. it will bring the digital haves closer to the digital have—nots. bring the digital haves closer to the digital have-nots. a lot of the teams are ready you started. we are expecting a lot of the teams to bring their own setup. but we want to use — at least offer the same to all teams. in the fifa system, the data analyst will have access to a tax on lap, allowing them to add drawing to a live video field. stills can then be sent to the dugout. with goalline technology, videos to referees, and now disco at this year's world cup looks to be the most advanced ever. with the technology, what is the risk of losing the rural spirit of the game? the role of the data analyst is simply to support the coaching decisions. i think there are some key insights they can be found in the data. but the football will a lwa ys the data. but the footba

in other words, we have to secure the data and use the data to benefit our customers because that's reallyuss model. without it there is no microsoft and no microsoft products. so to me, making sure that we live up to the high standards and expectations of our customers as well as the regulations that are already in place. take gdpr. in fact, we are taking the core tenants of gdpr and making it true for consumers and customers all over the world with our services. that's the approach we are going to take going forward. emily: you made big buys, including linkedin. i was speaking with your chairman saying he was more supportive of big buys that are as successful as linkedin. will we see you make more buys like this? netflix, there's perennial explanation of microsoft's interest there given it's the only big tech giant without a huge foot hold in the streaming space. satya: our first focus is organic growth and investments. r&d andig spenders in that is the core. we are always going to look for secular growth markets whether it's mindcraft or linkedin or now github. we look for things where

open standards the other people can look out and validate your data and cross-check your data. then it becomes much more reliable but if you're doing something on your own and not releasing that information only only talking about your results or something then who knows what went into it so the open data aspect of citizen science is crucially important. at the braman be project scientists are focused on the analysis of the data they hope to receive help from beekeepers from all over europe. there are thousands of citizen science projects worldwide that depend on data collection to get as many people involved as possible there are very few hurdles to getting started mostly all you need is a computer and passion but not every type of project is equally suited for everyone something that is more locally focused or. you know tangible to a person probably plays more than something like you know space travel or something so i think that really the limits are that you know the appeal is is what you can reach out and touch and what's around you and what directly impacts your life. the

i went on-line for some data. we ha we have over 4600 staff members, yet only 1,000 people have been trained in safety care, and only 310 have been trained in implicit bias. that's 6.7% of our staff, and we wonder why 40% of the office discipline referrals are for disruption? direct correlation. we need to do more around implicit bias. >> so my name is allison collins, and i'm so glad that lee said that because i'm looking at the same numbers, and i guess i just want to add onto that. at my daughter's school, there was a lot of -- i think at the school last year, we were the highest in suspensions for african american students in the district. we had several lockdowns, and we also had the police called on students to take them out of the school. this year has been a complete turnaround because we have a principal who's worked very hard to build relationships with staff and families and has also supported the staff in coming up with behavior plans that are consistent across the school. so it's been basically night

data protection is a fundamental right.f they don't match their provisions and regulations, they could be fined. there will be procedures before. vonnie: it sounds a little lenient. what do you make of the idea that mark zuckerberg at one point testified in europe and then was on stage with emmanuel macron -- are countries in europe more concerned about the jobs these companies are providing than protecting the consumer's data? jelinek: the government has made a strong commitment to the fundamental rights of data protection because they were the negotiators with the parliament and the commission. there's a strong commitment to the companies. uniquepanies will have selling propositions compared to their competitors. mark: how do you juggle the competing and conflicting interests of europe's data protection agencies? jelinek: the conflicting interest that might occur are going to be resolved by a special mechanism within the board. mark: can you explain how that mechanism works? i read already there are rumblings of this agre

data. we have to figure this out. the 600,000 pieces you are tracking today could be tens of millions or billions shortly. >> attribution is critical and some of the new technologies that are being developed right now could help us to attribute a piece of orbital debris that caused a loss of life to a specific nation or company. that is a challenge going forward for sure. >> thank you mr. chairman. >> i will now ask if you have any questions. >> thank you mr. chairman. thank you witnesses for being here. secretary ross, i would like to start with you, if you don't mind. i would like to start i stating my support for the president's space policy directive 3. and while i'm at it, i support directive 1 and directive 2 as well. we have a responsibility to protect the taxpayer. government spending and bureaucracy is a consider -- serious concern. however, not improving the space traffic management framework is on except double. what steps will be taken to protect against unnecessary spending and

their personal data.trengthen partnerships with academia. we partner with nyu to support their masters program. we recognize that these are our initial steps. we'll need to continually evaluate our technology and how we're doing and get more precise and build new approaches as new techniques emerge. we're counting on new approaches. in fact we're challenging the industry, it can be a giant company, academic setter, brand-new start up that just moved into their garage to build something, right? we're challenging the industry to build solutions that can scale to our great city and help new yorkers defend themselves online and fully protect their privacy. new york city believes that a strong, safe, fair and prosperous city depends on securing the digital space where so much personal and economic activity happens. we hope the new york city secure initiative will spark a discourse on how governments can take action to improve the cyber security for the public. all of us are deeply honored to bend the city's in

and censor data we have in the data center and around there into the system.thought that was the magic way, get it all in there and it's going to show us what we need to know. and what we found is, a, we kind of choked the system. it can't keep up with all of that. and b, there's so many things going on as you're trying to tune that, it's almost overwhelming in getting the amount of things down to make it valuable. so, people are becoming numb to different things, and it's just not being effective. and so, we've backed up on that to actually go after specific data types based off of, what do we want to do, what are we trying to solve here. let's bring in one data type, tune it, get the correlations and alerts we need first, then go on and add that next piece. second issue is the ongoing maintenance and tuning. so you know, the operations guys never leave anything alone in the data center. there's always some new system getting turned on, there's some updates going on, os updates or things moving back and forth between firewalls. and so, when you talk about kee

why sink a data centre? it's easier to cool data centres in water than on land. cool the data centre than we otherwise would. it is kind of a crazy experiment, but that's why we do research — to try these things and sort of push us to new places. and, if this is successful, maybe we'll have the ability to deploy data centres anywhere in the world very rapidly, perhaps as little as 90 days. this is the exact point where the data centre is going to be sunk and on the ocean floor is a cable bringing power to it, and then taking its data ashore and connecting it to the wider internet. a day on, with conditions in this sometimes rough stretch of sea very calm, the painstaking operation to put the data centre on the ocean floor is under way. 0rkney was chosen partly because it is a centre for pioneering renewable energy research, generating electricity from wind and wave power. 0rkney's produced over 120% of its electricity needs from renewables every year since 2012. so this is a place that's basically running on renewables, and so bringing a project like this to orkn

that been a scandal it's not a good signal to let there be unity get data from six not keep in touch with all our news twenty four seven r.t. don't call me and i showed you a bullet in here not international coming out we call him brain but thirty five minutes time for another for me kevin zero in a very good afternoon. when the makers manufacture consent instead of public wealth. when the room in closest to protect themselves. in the final larry go round the sun be the one percent told. to ignore middle of the room signal. for the real news is. the way. i've been saying the numbers mean something they matter the u.s. has over one trillion dollars in debt more than ten white collar crimes happen each day. eighty five percent of global wealth you longs to be ultra rich eight point six percent market saw thirty percent i just want to secure some with four hundred to five hundred three per circuit first second and bitcoin rose to twenty thousand dollars. china is building a two point one billion dollar a i industrial park but don't let the numbers overwhelm. the only number you need to

a lot of the teams already use data.o bring their own setup. but we want to at least offer the same to all of the teams. in fifa's system, the teams' data analysts will have access to a tactical app, enabling them to add drawings over a live video feed. stills can then be sent to the dugout. with goalline technology, video assistant referees, and now etps, this year's world cup looks set to be the most technologically advanced ever. but with the technology, how big is the risk that we lose the real spirit of the game? the role of the data analyst is simplyjust to again support those coaching decisions. i think there are some key insights they can be found in the data. but the football will always be those 22 players on the pitch, and the coach is always going to be the person making the decisions. at the end, what technology really is for is only to support. but at the end, you're a big football fan, i'm a big football fan, at the end, the 22 players plus the referee, they are the main actors. 90 minutes, maybe penalty s

law forcement to request data that u.s. companies may have, respecting and going through a laws of oking at the the countries, having the ability for those countries to object to that request, the ability for companies receiving the data request to also object , also setting a rulehat allows u.s law enforcement to obtain data that a company may be holding about u.s. persons, where that data may reside. there's a current court decision. it's unfortunate that this doesn't happen, but there is a current legal decision that requires more for content that law enforcement continues to follow. at the same time, the law allows for the u.s. government to negotiate treaties with other countries, like the united kingdom, for instance, to allow for those foreign countries to seek data about their citizens that u.s. companies may be holding, provided that certain things are met, including that they have a good human rights record, there's some congressional oversight, that they not seek information about u.s. citizens. all of this is ki

of data and we work -- you keep the data secure and use the data with permission from users.hat's been our corporate support, it's how we work with our customers and our platform. >> would you, like tim cook, differentiate yourself from the likes of facebook and while others may use personal data that that's something that microsoft has never done >> the thing is, for me -- one of the things i find strange about tech company ceos being asked about other tech companies, it's just not relevant to me what i want to stay true to is what is it that we, at microsoft, are in business for and how do we make progress that live up to the promise to our customers? trust is everything to us, whether it is our cloud services, our tools, our client platforms. to that point, absolutely, we'll differentiate on that point. and other companies may have different business models and different propositions and they'll have to make their own choices. >> on m & a, this is a sizeable purchase any more on the horizon? should we expect you to slow down to integrate this your stock is up on this news.