Welcome to Tech Talk. Bye.

the four companies that you mentioned, Facebook, YouTube, Twitter, and Microsoft announced that they would begin sharing hashes or kind of digital fingerprints of images and videos that are related to terrorist propaganda in some way. They've described in their announcement that this is the most extreme or egregious kinds of propaganda, but it's it's content that violates one of the company's terms of service against either support for terrorism or advocacy of terrorism, possibly depictions of violent content or or hateful speech, based on one of the company's assessment that, an image or video violates their terms of service, they'll contribute a hash of that image to a centralized database so that the other companies can scan and identify whether that same image or video has been uploaded to their services. So if Facebook sees a new post or a new video uploaded by ISIS, they will be able to sort of alert the other companies that this video is out there so that Twitter or YouTube can more quickly find it and also take it down.

illegal content? It's It's a good question. So the announcement from the companies was very focused on this being a voluntary process where the companies are each maintaining their own independent definitions of the content that they don't want to host on their own platforms, and it's going to be voluntary for any of the other platforms. And it's going to be voluntary for any of the other participating companies to decide whether they agree and want to take down, poster video from their own site. So if there's a gap between what Twitter considers off limits for its platform and what YouTube considers off limits for its platform, the companies may not come to the same decision. But, of course, taking a step back, this is all happening in an environment of incredible pressure from governments first and foremost on these very companies to do more to take down terrorist propaganda. Whether that's illegal, whether this we're talking about illegal speech actually really depends on the jurisdiction that you're in. Many European countries have laws against glorification or promotion of terrorism and advocacy of terrorist acts. In The US, we have laws against imminent incitement to violence violent action, which could cover some forms of advocacy to commit terrorist acts, but we don't have a law generally against, kind of propaganda that's generally supportive of a terrorist organization. So the question of whether the speech gets that gets swept up in this database, whether that's illegal, could depend on the jurisdiction.

based on terms of service removed removed across all of them and just disappear. Yeah. I think that's a real risk of this kind of system. Right now, the companies are very clear that they intend to each do individualized determinations. But I think you will end up seeing, that bring additional scrutiny on the companies for what those decisions are. You know, if one of these companies decides to take this material down, I'm sure there will be a lot of scrutiny if their partner companies decide to leave it up. Right. If one of your peers in the industry has said this is extreme and egregious content, that's gonna take some explaining to do about why you think it was a newsworthy event, why you think the context it's posted on in on your own platform really merits it staying up. I hope they do that. I hope there is really this nuance in how they kind of implement this, these shared hashes. But I think the risk is that that takes a lot of time. That takes a lot of resources and, could draw a lot of scrutiny on how they make those decisions. They also mention in the announcement about this program that they're, may in the future, open it up to other companies to participate. And, you know, if you think particularly for smaller companies, if you are a, you know, four person or even a 40 person little content hosting platform that has nowhere near the number of staff or the time and resources that the big companies do to do content moderation, the idea that there's a you know, the ability to quickly and easily take down, you know, some of the worst, terrorist propaganda or whatever ends up in this database could be very appealing. So I I wouldn't be surprised to see different kinds of companies take different approaches and and for some to adopt to just automatic takedown.

I think it depends on what impact you're you're looking to have. Right? Is is the aim to stop the spread of this particular video or this particular message from ISIS might be very, very effective at that. Right? It might be, and that's where some of the free speech concerns come in. This sort of system might be almost too effective in ensuring that certain images or videos cannot be posted on any of the the major platforms. You can think of scenarios where that's a really good good thing, but you can also think of ways that that creates real problems for journalists trying to report on these very images or researchers trying to understand, you know, how are terrorist organizations using social media to get their messages out? If your goal is to stop is to prevent the radicalization of individuals or to prevent the commission of terrorist acts, I think it's much less clear that there's a real connection between, the images and videos that organizations post publicly on social media and the ultimate, you know, legitimate aim of government of committing a terrorist act. There's a lot of research out there on how or increasingly, there's more research on how the radicalization process actually happens, and it seems like much more of that is focused on, recruiters grooming individuals and creating, you know, one to one relationships, lots of complicated factors, including the person's offline life and contacts and experiences that kind of the role of a video or an image in all of this seems very small in compared to the many other factors that actually lead someone to decide to commit a a violent act.

No. I so I think the one of the real risks that this sort of setup creates is that it is a centralized point for keeping material out of circulation in the public. And, right, everybody can probably think of certain material that they think should be in a database like that. What's not clear is how well the companies are gonna be able to keep this thing narrowly focused on only, you know, the most extreme and egregious terrorist propaganda. If it was a database only of, you know, videos and depictions of people being beheaded, that would there's still a strong argument that there's a lot of important reasons to be able to access that content, to be able to study it and review it, but that might be something that a lot of people could sort of get behind. I don't know that how possible it's gonna be for this to this database to remain that narrowly focused. Will it start being expanded whether on the company's own initiative or I think much more likely due to pressure from governments Sure. To cover a whole range of different kinds of speech. Kind of. Yeah. Exactly. Right. What counts as egregious propaganda on behalf of which terrorist organizations. Yeah. How far can that line move? I think it's really hard to tell without a very clear bright line rule about what content gets included in this database and also a lot of transparency and accountability built into the process so that there can be some kind of independent assessment of what's actually being registered in this database, you know, what is being kept out of public distribution on these major platforms, because people are gonna have a lot of questions about that. Right? Like this database existing is going to be pointed to for good, you know, for accurate

comes after several years of pressure from governments on these companies to do more to to keep this kind of content off of their services. We've seen the creation and expansion of these programs called Internet referral units in Europe, where governments refer content to companies for the companies to judge under their own terms of service whether that material, should stay up or come down. So it's this blurring of the line between government action on the one hand trying to get content removed from the web, but the putting all of the kind of onus and the responsibility for making that judgment and carrying it out on the private company. In this in today, the announcement from last week, with this hash database, I think you've got, you know, you've got the second part of that in place. You've got the coordinated way for companies to share information and assess assess this particular kind of content. I think it is only a matter of time before you see governments pick up the other side of it, right, and start pressuring the companies, whether informally or through court orders to put certain material into the database Yeah. Or to, you know, treat the fact that materials in the database as notice to the companies to all of the companies involved that there's illegal speech on their platforms. I think this is a big risk because as you mentioned earlier, you know, there's there's no judge involved in this process. Right? There's this is presented by the companies very much as a voluntary and independent assessment against their own terms of service. But I am not, sure how how to keep governments from trying to co opt this database and turn it into something where they issue reports of material they feel is illegal, and and use it as a much more formal hook for the companies all to keep this material off of their platforms.

So we've talked about a few of them already. I think having a very clear and very narrow definition of what material can go into this database is gonna be very important. I think it's very understandable that each company wants to maintain their own independent terms of service and their own evaluation of that material, but without a clear standard that, you know, make it more restrictive than what actually violates each company's terms, put a very high bar to what can be shared in this much more powerful kind of system, so that it's much easier to say, no. That image or that video can't be added because it doesn't meet that bar. That's gonna help guard against mission creep, and that's going to be a you know, if that can be a clear, and consistent standard that they continue to apply, that could be the best argument against the very many pressures from whether it's from government or civil society about adding other kinds of horrible content to this database. I think transparency obviously is going to be a very important piece. Understanding kind of what is the effect of this database. Does it lead to, you know, hundreds of thousands of images being identified, or is it really a very small number? What is the how are the responses of the companies actually bearing out? You know, does everybody decide to take down the material that they see? Is there room for one platform to make different decisions in different circumstances, right, where when a image is posted organization, it is considered to violate their policy against advocacy for terrorism. But when that same image is posted by a journalist, it stays up. That's the kind of nuanced response that we'd wanna see. But without kind of reporting from the companies on how it's carrying through, we're not gonna be able to tell that. And then on how it's carrying through, we're not gonna be able to tell that. And then finally, I think having some kind of independent assessment of this, process is gonna be really important to, allaying the concerns that I think governments, civil society are gonna have about what, you know, what is this thing and how is it operating. There's a chance for this all to seem very much like a shadowy collusion amongst some of the major tech companies in a time when lots of people a So having some kind of independent assessment that can be a guarantee that the material contained in this database really is only material that meets that clear bright line and hasn't turned into just the kind of catchall for copyright

Should've asked first. Me too. Joins us on Tech Talk today. Welcome, Amy. Sorry about getting your name wrong. That's okay. Kick it off. So first, for people not familiar with your organization, tell us a bit about the data quality campaign. Great. Thank you so much for having me here. We're it's a pleasure, and we're so glad since Nuala is on our board. It's a great, great tribute to have here. And so we really appreciate our partnership with CDT. So thank you so much for talking about this important issue. So data quality campaign is a nonprofit advocacy organization that's focused on changing the role of information in education. We believe passionately in the power of information and of data to change education and to change kids' lives. And we believe that our nation will never reach its education goals if we don't really focus on the power of data to empower better decision making, to provide greater information on returns on investments, about what's working in education? And how do we also make sure that those closest to children have the information that they need to make the right decisions, to take the right actions, to get the outcomes that our children need and that our country needs as well to

data could be used to harm and hurt and it can be used to help and to really, improve things. And at the data quality campaign, we've really spent the last eleven years since we were founded to really change the conversation about the role of information, the role of data in education. And to change data from being seen as a hammer that's used to hurt people, hit them over the head, to really using it as a flashlight, as a tool of empowerment, a tool to shine a light on what's working, a tool to really show what's working, what's not working, and how do we use that information

What would you say your top level message is for them? So I think our top level level message is what our message is for everyone that we talk to. And our big idea at DQC is, you know, when when people have the right information, people that are working, whether they're parents, whether they're students themselves, whether they are teachers, educators, policymakers, voters, citizens, when people have the information they need to make the decisions that they need to about education, students and children excel. And And that's what we're all about. How do we make sure and we want this administration to focus on the power that they have and the and the roles and responsibilities to make sure that anyone that has a stake in education has the data, has the information that they need to make the best decisions possible to help every single child in this country succeed. And that's our recommendation that we've been making with state policymakers, that we've been talking with federal policymakers, and we feel there's a real opportunity with this change in administration, at the federal level to reinforce this message and to also work with congress on this, not just the administration,

The first one is to ensure that families have access to their own child's data. Is that not the case now? It's not actually. You know, we, for the past eleven years have surveyed states about about the status of data collection and data use and data empowerment, who sees what at what point. And our last survey that was done in 2014 showed that only 17 states at this point are able to guarantee that parents have their own child's information. And We're not just talking about test scores. We're talking about meaningful information that can follow a child over time so that you don't just know how your child's doing today, but how does that look on the story? So really the video, not just the snapshot of how well a child's doing. And when you think about it, you're right. That's crazy that we are not even providing parents and students themselves the information that they need to be a great champion for their child, to be an advocate for their child, and just to make sure that they're making decisions based on information rather than anecdote or hunch or what they think is the right thing. But they actually can have some data to really back up their role as their parents' greatest as their child's greatest champion. And you would think that having access to the sort of information that is, you know, kind of creating the educational path for their child would make them feel a little bit more comfortable then about data being used in education if they knew what it was and weren't worried that, you know, either their their child was being held back or whatever the case may be then. And Brian, that's an absolutely critical point. Part of the reason I believe that we've been seeing this backlash against both data and backlash against testing is that most people and most parents when you say the word data in education, you think about a test. Yeah. You think about a high stakes state test that people feel like they've not gotten any value out of. They don't know how that information's and how that test is going to help their child. They rarely see the information in a timely basis and it's really has not been explained to people about the what's in it for me. Why does this actually matter for me as a parent and how is this gonna help my child? So as a result, because they don't see value in it and they don't know how it's being used, and there could be some mistrust there, guess what? Then they are gonna say, I don't need that. I don't need this data. And I'm kind of more worried about it. And this back to your earlier question, that then people say, well, wait a second. This sounds more scary than good. And, you know, when people make that decision about, is this gonna help my child, or is it gonna hurt my child? People will always, fall on the side of making sure their child is safe and secure first. Absolutely. And so what we think with this first recommendation is that a very important foundational step, is to make sure that every parent and every student, him or herself, has their own information. Because at that point, we have a better understanding of what's happening with my child, and it changes the entire conversation. And to be blunt, it takes the boogeyman out of some things because if you actually see what people have collected on your child, it changes the conversation and it puts aside some of that fear and it builds trust through that transparency. Great point. So your next one was, actually, it seems to piggyback right off that increased transparency and public accountability. Obviously, you've touched on transparency. What about the public accountability side of it? So this opportunity and we, in education, we had a landmark piece of legislation reauthorized, last December, the Every Student Succeeds Act. And this is a reauthorization of the much vilified no child left behind act. Mhmm. And one of the one of the things that Every Student Succeeds Act does is it builds on what I think is the greatest legacy of No Child Left Behind, which was providing public transparency about how well students and groups of students are doing. The every student succeeds act really builds on this idea of public reporting of of progress of schools. And what's different and really important and why we think this administration should double down on this is that there's an opportunity to not just opportunity to not just have that that transparency and that accountability be focused on the fed federal government and states being accountability to the federal government, but more about how do we make sure that the public at large has the information that they need and that they deserve. And to be blunt, they haven't been getting about how well their neighborhood schools are doing, how well are our tax dollars being spent and are we getting the results that we need and are we serving every single child that we need to serve in this country? And so the opportunity here is to build on what this new reauthorization has done and actually improve upon it. So for example, one of the things that's now required in every, every state in the country is to report back on feedback information. How well do the graduates of this school do when they go into, when they leave our doors and they go into post secondary education and training? And you know how to report back, what percentage of students of your graduates from your high school go on to higher education, and that's really important. That's what parents wanna know. They wanna know what happens to kids once they leave these doors. Is my kid gonna be okay? And yet we think that there's so much further that can be gone of what parents really wanna know and what teachers and principals really wanna know is, well, of that percentage of kids that went on to higher education, how many of them dropped out? How many actually were able to get a job? How many needed remediation to retake courses that we told them that they did okay and in high school? And so that level of information. And and the bottom line is this point about transparency and public accountability, we need to make sure that the public at large has the information that they need and deserve, not just our policymakers, but the public about how well our schools

knowledge, access, and the tools they need. What how can data help out teachers? So one of the things that we know about our teachers is that they are starved for better information. They know themselves that data is a is a tool that they can use to make sure that they are personalizing education and tailoring education to every child in their classroom. That they know that it's a tool that's gonna help them make their jobs easier, that especially when combined with technology, they have the ability to tailor just in time and get back instant feedback so they know how to change the lesson plan, not just for next week, but to make sure that every child's on track for mastering the concept they're teaching. Really exciting things. But what we also know from lots of surveys and listening to, teachers, they don't know how to use this data. And so we have a real lack of what we call data literacy in this country. There are very few schools of education that are are offering courses. There are very few states, that actually require that, that teachers, that perspective teachers have to be data literate, that they have to know how to access and use and protect data in a responsible way. And so we think this is a real opportunity for this administration to underscore the important role that teachers play in making sure that data is used as a tool for good. It's It's used as a tool for of empowerment. It's used as a tool for personalization. Because what we also know is that information and data by itself doesn't change anything. Right. That unless people trust that information and unless they find value in it and they know how to use it to help meet their goals, then it's all for naught. And so we think this people side of the equation, the people side of the data equation really, really matters. Yeah. And teachers definitely work very, very hard. So without kind of that literacy, trying to figure out how to use the data, that's just more that they have to do in terms of their day. So finding a way to make sure that data is something that makes their their lives easier, you know, improves outcomes. It also makes their job easier and, you know, makes them more effective, but not adds another layer of burden. If I can add to that, I think one of the key things is that it can't just be another thing as you're saying that adds on to the long list of what we're asking our teachers to do. And this is also when we talk at DQC about we need to improve and and address the conditions and the capacity and the culture around data use that without also changing how we think about time, how we organize the day, how we're providing opportunities for teachers to talk with one another about students using data. Then again it's just an add on and teachers end up resenting it as being something that's hurting them rather than something that's helping them be successful

one of the key pieces about building trust is having people understand why you're collecting a piece of what data you're collecting, why you're collecting it, how it's gonna help somebody, and what are you doing to make sure that this data is used only for the purpose for which it's collected, and what are you doing to ensure that it's kept safe, secure, confidential, and private? And I would argue without doing all of those things, we will never ever ever win the public trust and especially the trust of parents. And that's what's critical. As I keep saying, no one is gonna use information that they don't trust. And one of the things we're finding is that there are lots and lots of questions out there about, well, what is collected on my child? Why are we collecting it? How is this gonna help somebody? And like in every sector, we have lots of data that has been collected. It's been on the books for ten, twenty, thirty, forty years because somebody wrote something into a piece of legislation, you know, here in DC on the hill or in the state capitol somewhere and people aren't using that information anymore. And so one of our recommendations is to do an annual audit both here in congress, the administration should do this, as well as at the state level to ensure that we are reviewing every single year what are we asking people, what are we mandating we have to collect. And if somebody can't prove about how that piece of information is going to help a child succeed, we should sunset that collection. That would go a long way to building trust and making sure that people that someone's mining the store, someone is thinking about what we actually need and we're not just collecting things willy nilly, but that there's a very deliberate plan that we only collect what we need and that we can defend is going to help children succeed. And there's a whole lot of recommendations that we've written in here, about what we need to do about data governance, what we need to do about ensuring that we're using best practices from not just within education, but across the entire world. And this is the work that you all do so well of identifying those best practices about security and privacy and making sure that the education sector becomes a leader in this, because that's what trust is all about. And we have to make sure that the data conversations are not ones built in fear, and myths and a sense of people not knowing what's happening with their children's information and instead double down on making sure that people understand why this data matters, why it's going to help their own individual child, and how this information is being, not just collected, responsibly. And that also this faith and this trust that everybody who is working with this data from teachers to administrators, are know how to use it responsibly and ethically and morally.

around data. So we have a lot of holes and clunky systems, I guess, is the technical word, and about our inabilities for data to follow and work for students on behalf of students. And that's what we're all about at DQC. How do we make sure that data is working for students? How do we make sure that it's benefiting students? And right now, we have lots of things that are working. And so, you know, we always believe at DQC start with the questions that people have. And right now, we we can answer a whole lot more questions than we could ten years ago before we had the development of data infrastructure at the state level. But we still have lots of holes in those systems. Right now, we cannot follow the information cannot follow a student. So when they leave, once they go across the state line for learning. So thinking about the opportunities to do online learning, or to take a higher education class in Massachusetts and you live in Minnesota. That information right now is really hard to follow. How do you make sure that that information about how well a student's doing on an online class actually gets to be part of that record? How do you make sure that we're able to connect up a higher ed record, you know, the records I talked about earlier, that post secondary. Right now, we don't have the ability to have data that feedback data to follow a student if they go to a higher education institution that's different than where their high school was. Oh, interesting. Which is crazy. Right? That when we talk about competitiveness and we talk about how do we make sure we're preparing every student for success in life and in a knowledge economy, we actually don't know how well we're doing because it's so hard to actually connect that information. So, you know, feedback information really matters. And if we want our k 12 systems to improve, they need that feedback information. And yet right now, that's the kind of infrastructure that's lacking and that we think there's an opportunity to build that information structure. And again, that needs to be driven not by the needs of data nerds. It needs to be built by the needs of what individuals need. You know, number one thing, what do parents and families wanna know when they are now starting to look at making the biggest investment in their life in higher education? You know, taking out thousands of thousands of dollars of loans. They wanna know what am I gonna get as a result of this? If my kid goes to this school, what percentage of kids graduate on time? What kind of salary should they get? What kind of jobs? Things like that. That information is really hard to come by And especially I wanna know, can you link it back to a high school and know kids that graduate from this high school, how successful are they in a in a higher education? What are the factors that made them successful?

So I personally hope so. I I believe that the success of this administration and the success of this country is dependent on how well we educate every child. And their, the power of the bully pulpit of the presidency and the federal government for setting the direction for the nation and for setting the tone of the conversations that we're going to have and what we're gonna prioritize matters a whole lot. And so I really do encourage the the Trump administration and new members of Congress to focus on the power of data to change lives and the power of data to ensure that that our school systems are serving, serving every student in this country and that we have an opportunity to make sure that we get the best, results for our taxpayer dollars, but then also that we take advantage of innovation, that this can help support choice. And we know that some of the things we're hearing about, the themes that will be important to the Trump administration such as international competitiveness, such as increased choice. Mhmm. Those things those conversations can't happen unless you have quality information. Talking about empowering parents. You can't empower parents without giving them the information they need to make the best decisions possible. So I sure hope that education is a priority of the Trump administration and of the next congress. And I really, really hope that the conversations around how we make sure data is a tool for good and that we make sure that we're using data to empower, and to inform better decision making at all levels is an absolute priority for this administration.