Please welcome back to the stage insa president suzanne hackenberg. [applause] suzanne i hope everyone has enjoyed dinner and is moving to desert. Before we resume, i have two quick announcements i would like to make well i have your attention. First, i am priest to announce the foundation today announced th please welcome back to the stage insa president suzanne hackenberg. [applause] suzanne i hope everyone has enjoyed dinner and is moving to desert. Before we resume, i have two quick announcements i would like to make well i have your attention. First, i am pleased to announce the foundation today announced the seven Scholarship Recipients we have built over the past three years. Cumulatively, these scholarships are valued at 50,000. The rewards make a difference in the lives of undergraduate and graduate students all willing to work here for the rest of their lives. [laughter] three of our recipients are here with us this evening. I cannot help but point out they are all women. [applause] brooke rubert is a master student at marymount university. Can you stand . [laughter] samira nooro is a master student at marymount university. [applause] emily robert is a dual Degree Program at penn state university. [applause] i really think our future is bright with these young ladies. If you have not had a chance to get to know them, i bet they are all pursuing careers, if not, doing internships in the Intelligence Community. Thank you for joining us this evening. Finally, the registration open ed today for the new i. T. Program on september 27. The fullday program focuses on the challenges and opportunities our Community Faces as we build, foster and advance gender inclusive and a Diverse Workforce. We will have a wide ranging panel. Plus, plenty of networking time which insa is often known for. , it is my pleasure to introduce our moderator tish long. Her accomplishments are almost too many to count by well anyway. Deputy director of the dia. Principal deputy under secretary of defense for intelligence and Deputy Director of Naval Intelligence chairwoman of the , insa foundation. We are thrilled to have our conversation this evening with dr. Dixon. Over to you two. Intelligence ad Deputy Director of Naval Intelligence mentored a. As well as a terminal a chairwoman of the insa foundation. We are thrilled to have our conversation this evening with dr. This dr. Dixon. Over to you two. Thank you. Let me just as mike and welcome. A special welcome to our Scholarship Recipients. We have the good portion, stacy and icom are to have two of the recipients at our table. Our future is bright. I want to recognize stacy dixons parents as well. I had the good fortune of meeting them. When was it . Dr. Dixon 2010. 20 years ago. Stacy mentioned her promotion to capitol hill. So glad you could be with us this evening. You sent something in your remarks about i probably did not know you were going to be the pd dni. Dr. Dixon that is true. I dont have a crystal ball, but like your mother, i knew you were good people and i knew you would go far. I am not the smartest person so i like to surround myself with smartest people. A customer came up to me earlier and said dr. Dixon is the smartest person i know. I said yes, i hired her. Even if you did go to the other tech school. Thank you again for being here. We have been asking you for a little while so glad we could finally aligned calendars. As you can see, and lot of folks wanted to your from you this you think. I would like to key off a couple things you set in your opening remark. You talked a little about ukraine and russias invasion of ukraine and the impact this is having from a global perspective. One thing you did not mention, that we heard from the Principal DeputyNational Security advisor a couple weeks ago and he discussed the policy on what he refers to as strategic downgrades. It is all about information sharing and strategic youth of intelligence. Can you talk about this . Not only the impact it has had on the hostilities and the war, because that is what it is, but how are you thinking of the release of intelligence go forward . What is this mean about information sharing with allies, etc. . Dr. Dixon you can imagine we have been getting a lot of questions about this going word but in terms of how and initiatives vision, how we came up with the initial position, allies and partners are seeing things the same way. How do you have the conversation and make a decision knowing you have give up your sources to share this information . In conversations with the administration, it is a policy driven thing the Intelligence Community responded to. They wanted to respond to many things on many levels and the only way to do this would be to downgrade information. How do you downgrade for your partners . And the broader definition of partners, given the number of partners involved. Nato plus. The next question is, can you downgrade this . So how do you balance this hunger you created and being able to share more information. There is a lot of risk that happened by a lot of people in the community. It goes back to those losing the most with information revealed. We dont want to give up the information we had today that will prevent us from having this in the future. Our teams have worked really hard to try to get the downgrading process to move more quickly because of the speed needed and also the amount of requests coming in. Is there a Technology Aspect that can help with the speed of downgrading . I know a lot of human interpretation has to happen. We get a lot of smartphones and Technology Helping. Dr. Dixon right now there is not a lot of Technology Helping but we see the future where we have to increase our technology. Not everything needs to be classified forever so getting to the point where you let Technology Help you because there are a lot of holding, hitting the timeline where you review them for classification. There is legislation requiring us to look at how and when reclassified. We know we need to do it internally if nothing more than to make the process helpful to ease for folks who do this daytoday. Certainly with regular processing and there are tools we can leverage. And in getting the information out to the American Public. It is important. I want to come back to technology. Get flipside of where we started with downgrading intelligence to share with partners. There is a life information a lot of information and open source. We sometimes joke it joke if it is not secret, it is not worth looking at. I know that is not the case. I know the stacey dixon i know the odnis office is looking at that. Are we going to have a another agency in the community . Dr. Dixon our goal is not to have another one. There are studies suggesting that is what we need. We think we can accomplish the goal without starting another agency. The problem is you need a good portion of people just to run the agency. If you are looking at, has you keep organization small, that stuff away way to do that. How do you create a tradecraft and practitioner all consistently trained around the community to do open source intelligence. That is where we are headed. We have had opensource enterprise in the dia has opensource enterprise as well. We need more consistent training between practitioners by would be the first to say the community writ large is a lot of value as opensource. Back in the day, it weakens the criticisms that we were not good enough or true. We dont talk about them as much for good reason but we are interested in what it can do for us. How do you make sure everyone working in similar ways is treating this just like intelligence where there is a tradecraft and standard . There is going to be a Small Organization that will help get things kick started. Then you hand it back to functional managers to deal with it. Why do you think the community has not been able to do everything you said . Dr. Dixon we had an opensource organization, the Community OpensourceProgram Office in the late 90s. We have been talking about this for a long time. How can industry help . Dr. Dixon your commented that if it is not from a sensitive source we control means it is less trustworthy there is a piece of this we have to deal with. However, have to figure out what we can use it or and how we can trust it . How do you build a providence as initial starting point and then use your sensitive and more scarce info for other things . This is one thing without having the tradecraft. There are people whose first primary mission is to start using open source. We do not train people like this. Opensource for First Enterprise use focus on translation. There is more than just the translation part so we need to bring that in. What are the best things we can get from it . We can learn from organizations to get great intelligence for their own disciplines and needs for open source material. We have a lot to learn and industry can be a part of it. We both know nga is a good job in using commercial industry. The state department does a lot from their diplomats. There is a lot out there that is being used now. I think this is an initiative where i certainly hope you will make real progress. And i think others do as well. So lets talk about technology because technology can help you there. And just across the board. Maybe start with positive aspects. How is Artificial Intelligence being applied now . What are some shortfalls . Maybe initial Lessons Learned . Again, how can industry help . Dr. Dixon when you look at our community and how much data we are sitting on right now, your initial thought is this would be perfect for ai. Data that was collected for other reasons and is not tagged for ai use. There is a barrier to be able to use what we have but with data officers around the community, we are thinking about how do we make sure the data we have from this point on is ai ready . We have catch up to do for other data and then it is a matter of, do you call all resources into tagging . There is a love of cost that goes along with it. We cannot do the same thing china would do and have armies of people that do nothing but tag data. That is not costeffective for our community. This is one of the first run first lessons that ai was not ready for. Bringing in people now who can write the necessary algorithms on that level two understand the on that level and can understand the code. That is a good thing. It is a workforce we are developing. It is something that, as you mentioned with the data strategy, being able to have the relevant data as a workforce is a goal. We are using Machine Learning in certain really good cases to get effect. It is not nearly as widespread as anyone knows it needs to be. I have seen more in automation first steps. How do you make sure you get strategic things over and over again. The use of Machine Learning is the next one. To really use it to help you make decisions. This will be a step in the future. While we are doing this, we have the opportunity to build on the Civil Liberties protections and ai protections we need. We are trying to do this in a smart way. Do not want anyone to judge us as not having taken heed to what we are hearing. Researchers talk about the biases. How do we make sure we are not doing that . Chair long of course, there is a lot left to do. The technology has since changed so rapidly. How are you thinking about the guardrails, the framework, the ethical use of i hate to have to quested this because i am not sure we know what true is that have to question this because i am not sure we know what the future is. Dr. Dixon it will be a wild for the most critical decisions. We are excited about the technology and where it can take us. Because it is so widespread, the barriers to entry are so much lower that many adversaries, individuals who never would have been able to use this technology for their own purposes, now have access. One great thing is this is the advent of Large Language Models and the splash they made and av maintenance how quickly they were adopted for good and bad. They are looking to government to allow us, or encourage us, to create regulation and legislation in a way that will help protect without harming business. Recognizing there are a lot of people that will use this quickly for bad purposes. How do you make sure guardrails are there and models are protected . There are worries. But expecting what you can do with these models and unclassified data later, there is a lot of excitement. We want to do this in a smart way, methodical, having thought through the what if. Chair long it says something when you have ceos of Major Tech Companies citing letters to congress that we need to think about companies signing letters to congress that we need to think about the things happening here. Dr. Dixon there are certain fields where it has taken a long time for government and industry to work together. Cybersecurity for example. It has taken a a while to get to where there is trust between the organizations. We need to keep this. It is safe by everyone talking to each other and making sure if it happens to one organization, it is not just them. Shared protection is important. Chair long we are connected. There are cards on all the tables if anyone has questions. I have plenty but i do like to ask questions from the audience. I will merge one i received from the audience earlier with one i had, golf teeing off something you said. That the data was not ready for machine algorithms. He talked about new strategy. Can you tell us about this and what nearterm and midterm impacts this will have a Big Community . Dr. Dixon that is a good question in terms of timeline. Part of it is helping us think of ourselves as a data organization. We are intelligence, but the thing we are focusing on and have is data. If we continue to change the mindset of people to think about data first in terms of how you are building systems, it is to build data. It is not just intelligence being done in its own separate file. You want to use these things together to get effect. I dont know exactly what they will be able to achieve by having vision that bringing the data together earlier in the system and not waiting until the end will be very powerful. But we have to build systems. Today, mostly every organization is to be its own thing. So you have great Repository Data which is different from the other great Repository Data. There is more potential in the future if we build our systems from the mechanics. We have to come up with Data Management plans. How do you think about data and to and from the moment you collect it to how you will use it to how you will dispose of it . How do you make sure you have analytics on the scale or level you need it and at the speed you need it . These are things we need help with. Chair long to tie it back to something earlier is the declassification. Thinking about how you tag the data upfront for 25 years, down the road when you have mandatory class of mandatory declassification. Part of the audience question was do you see the introduction of multicloud helping . Dr. Dixon absolutely. I am excited that we have several cloud providers even at the ts level. They just added another one, i think. We are excited to have more in the family. It will be powerful for our community to have access. It helps with resiliency. You can open up a lot of opportunity for it. Chair long well said. A number of questions here. Lets see where we want to pivot next. This is actually to both of us. Dr. Dixon ok. Chair long i have never gotten a question for both of us. I was going to ask you, one of the things both you and dni haynes have talked about is developing the workforce. Yes, it is about the data. At the end of the day, it is about the people. I really like this question. As two strong, intelligent women , how do you help diversify the icy . Both government and industry. And how do you ensure youth are engaged and interested in the icy . The ic . Dr. Dixon one thing we have been doing is capturing the statistics of what our organization looks like. We are doing this now in a much more detailed and automated way, then showing results to the workforce. We have data dashboards that show the demographic makeup of our agency. And let us know what we look like compared to the American Public. Part of it is knowing where you are starting so you know where you need to head. With that, you can create great strategies of recruiting at schools that maybe dont see anyone from the Intelligence Agency come to them. Whenever i go out traveling, i try to tack on college and high school visits. Just to see students hearing about the Intelligence Community for the first time. Having someone say, there is a place for you here. If you want to come serve your country, or tech National Security, and get into interesting inns over the course of your career, not only are there people who look like you but those who are successful. Knowing who we are into we need to commute to really look like america. We do not always know the best way to talk to different communities so recruiter training and helping people understand how different communities hear the information we are providing. How to make sure we are not using triggering words or words that will turn off the community. We try to be smart about ways we are advertising but we have to hold ourselves accountable at the end of the day. We are changing to a system where we will work more with opm to collect more Demographic Data that we already collect, to find out people are not making it through the process. We need to be able to figure out this so we