Cybersecurity

Security Challenges in Education Today

Security Challenges in Education Today

This is an audio conversation from the CIO Talk Network podcast. The timestamps in the chapters and transcript below refer to the audio recording.

About this conversation

A CIO Talk Network conversation with David Sherry (Brown University), Joshua Beeman (University of Pennsylvania), hosted by Sanjog Aul.

What the conversation established

The conversation established that securing a university differs from securing a business because the mission is teaching, research, and service, which depends on openness and collaboration. Joshua Beeman of the University of Pennsylvania described a decentralized setup of 12 schools where personal responsibility outweighs central control, since any user can start an Amazon cluster or copy data onto a personal device without IT’s knowledge. David Sherry of Brown University sorted policy into what people can, may, and must do, putting HIPAA and FERPA data in the mandatory tier, and said higher education felt like the wild west after his years in financial services. Both argued that education, clear policy, and incident response beat threats and fear. Beeman cited Penn’s Security and Privacy Impact Assessment, adopted by every school and center, while Sherry said a security budget needs strong business plans and a three to five year vision, not fear, uncertainty, and doubt.

Host Sanjog Aul talks with David Sherry, Chief Information Security Officer at Brown University, and Joshua Beeman, Chief Information Security Officer at the University of Pennsylvania, about the security challenges facing higher education. The two CISOs explain why open, collaborative campuses are harder to secure than corporations, and how decentralization, bring your own device, and self-service cloud shift risk onto individual users. They compare how policy, education, and incident response work in a university versus a business, and discuss budgets, frameworks, and leadership. Sherry and Beeman draw on examples from Brown and Penn, including Penn’s 12 schools and its Security and Privacy Impact Assessment.

Contributors

  • David Sherry, Chief Information Security Officer, Brown University
  • Joshua Beeman, Chief Information Security Officer, University of Pennsylvania

Host: Sanjog Aul, Founder and Host, CIO Talk Network

On this page: Key takeaways · Chapters · Explore more · Transcript

Key takeaways

  • David Sherry, Chief Information Security Officer at Brown University, said moving academic life online raises not only personal safety risks but also academic dishonesty, plagiarism, and threats to intellectual property and research. He noted that the cloud, mobile devices, and bring your own device have opened security gaps that did not exist as recently as five years earlier.
  • Joshua Beeman, Chief Information Security Officer at the University of Pennsylvania, said the core problems of poor judgment, cheating, and bullying are not new, but technology has made their effects faster, wider, and more permanent. He contrasted his own college years, when a mistake was forgotten by the next morning, with a student today whose unflattering photo can go viral the same night.
  • Joshua Beeman of the University of Pennsylvania described a decentralized institution of 12 schools, including the Wharton Business School and the Annenberg School, each responsible for its own IT needs. He argued that personal responsibility now outweighs centralized control, because a user can invoke an Amazon cluster or move large data sets onto a personal device without the IT department’s knowledge.
  • David Sherry of Brown University framed policy as three concentric circles of what people can do, may do, and must do, with mandates like protecting HIPAA data in the must circle. He said that after coming from financial services, where he could dictate rules, higher education felt like the wild west, which made balancing risk with the university’s openness both harder and more rewarding.
  • Joshua Beeman of the University of Pennsylvania said people do not respond to threats, and that education paired with detection and automated controls works better than authoritarian warnings. He pointed to Penn’s voluntary Security and Privacy Impact Assessment, now used by every school and center, as evidence that strong data beats fear, uncertainty, and doubt.
  • David Sherry of Brown University said security budgets require strong business plans and almost MBA-like thinking rather than fear, uncertainty, and doubt, because a project that prevents any incident can look like wasted money. He manages security as part of Brown’s overall risk management framework and plans against a three to five year vision of the security architecture.

Chapters

  • 0:00 Introduction: security challenges in education today
  • 2:24 Risks of putting campus life online
  • 6:26 Strategy and firefighting in a typical day
  • 8:17 A decentralized campus and personal responsibility
  • 16:38 Why mission shapes security policy
  • 19:06 What people can, may, and must do
  • 22:21 Education over threats
  • 31:23 Involving students in policy
  • 33:34 Innovation, social media, and risk
  • 40:20 The security team’s mission
  • 50:15 Justifying security spend without fear
  • 52:12 Measuring risk with data not opinion

Explore more

Transcript

Introduction: security challenges in education today 0:00

Sanjog Aul [00:00:26]:
Today’s topic is security challenges in education today. And our guests for today’s show are David Sherry, who is the CISO, which is chief information security officer with Brown University. Good morning, David. How are you?

Sanjog Aul [00:01:07]:
And we also have Joshua Beeman, who is the chief information security officer with the University of Pennsylvania. Good morning, Josh. How are you?

Sanjog Aul [00:01:24]:
So, while we all talk about weather all day long and it changes, but at the same time when we are talking about education, we want to ensure that whatever the platforms that we are building as an organization or as a business for profit or not for profit, in any case, the people who we are trying to strengthen, build a next generation workforce and leaders, we built a strong, secure environment where these people can go about getting the education and people who impart, they are also safe and secure. With all that said, David, let’s start with you. Across the board, there is actually a rising concern over campus crime and safety. And when we are talking about everything going online, that is electronic age that we have and the student interaction is also being preferred to be done online, what kind of risks that we are further increasing and are introducing?

Risks of putting campus life online 2:24

David Sherry [00:02:24]:
Very good question to start off with, Sanjog Aul. There’s obviously many facets of that that we can go into. There’s personal safety and personal dangers when students are interacting online and putting their personal lives online, for most of the world to see. But there’s also the risk to the academic process and to the institution itself. When we have things online, it introduces academic dishonesty and plagiarism and copyright a little bit easier than it was in the past. There’s also the concerns over say the loss of intellectual property or research protection. And, in the era of the cloud with mobile devices and bring your own device or as we say in higher education bring your own devices, and the students and faculties need to have data access anywhere, anytime on any device, it obviously opens up security loopholes and security challenges that we didn’t see as short as 5 years ago.

Sanjog Aul [00:03:20]:
Now, this is a question for you, Josh. Do you think this is a sudden shift because we are trying to go online, or is it basically while you’re trying to morph the business to remain sustainable and profitable, this is coming as not a fringe benefit, but as an add-on risk, which was not warranted?

Joshua Beeman [00:03:40]:
I think it’s important to point out that the security issues that David is describing and that you’re describing, these occur across the board, right? They’re not just happening in higher ed. But from my perspective, I don’t think that these broader problems are necessarily new, things like poor judgment or cheating or bullying. But the means by which they’re occurring have changed, and sometimes the problems happen so fast and with such widespread effects and permanence that the impact is a lot more profound than it’s been previously. When I was in college, 20+ years ago, a student could make mistakes, and they would probably be forgotten the next morning. Today, a student could find an unflattering photo or video of themselves go viral the same night, because of some lapse in judgment.

David Sherry [00:04:32]:
I was just going to add, another interesting shift is just the way we look at information overall. I’m going to tack on to what Josh said about when we were in college. I used to get my grades by visiting the professor’s office and looking up my social security number that was posted on his door. So, things have changed rapidly, and the way we look at protection and the way we look at information certainly has changed dramatically as well.

Sanjog Aul [00:04:55]:
Now we always take pride as information security officers or, in fact, anyone in technology to be proactive and have the capability to really sniff and sense any dangers that we might be coming across. But is it truly the case when it comes to security whether in-person or information-based security, Josh?

Sanjog Aul [00:05:23]:
What we’re seeing is, are we able to proactively sense the danger when it comes to the kind of security issues that we are seeing in education, especially when we are dealing with individuals who may not be as mature as an adult that you’d expect in a workforce?

Joshua Beeman [00:05:41]:
I think in general terms, yes, we can anticipate certain types of issues, whether that’s loss of judgment or fraud or theft, the same kinds of problems we see not connected with technology. But at the same time, individual attacks, those very specific nature of the technical attacks that we’re seeing are changing on a daily basis.

Sanjog Aul [00:06:04]:
David, do you think your typical day goes about fighting fires or plugging holes, which may have been introduced which surprised and/or shocked the management, in the areas of security, or you get some time to do some daydreaming where you can proactively think about what could happen next and also have the resources to plug those holes beforehand?

Strategy and firefighting in a typical day 6:26

David Sherry [00:06:26]:
Well, first of all, there’s no such thing as a typical day or an ordinary day. I guess my ordinary day is that I never know at one day to the next what my day is going to entail. And I think, Sanjog Aul, it’s a little bit of both. I think we have to spend some of our time in the strategic realm whether it’s business risk or risk assessments or working with the risk management committees assessing what’s going to happen with the upcoming technology or even business continuity, disaster recovery, whether we’re going to get a snowstorm in the next few days or whether we have a hurricane coming up the coast. So there’s always that strategic thinking. Certainly, there’s tactical roles and responsibilities that have to go on on a daily basis. Some of it includes firefighting, or log management solutions are always telling us things that we need to look into, maybe a compromised machine, maybe a spam outbreak. We all certainly have that. Sitting back and thinking that that’s kind of a luxury during the regular normal business day. I have to do that early in the morning or later at night or in conversations with my friend Josh and other peers, we kind of do our wishful thinking together to find out who’s doing what right, who’s doing things that may be able to be improved upon, and how we can work together as a security profession and a security discipline to make the next strategic moves.

Sanjog Aul [00:07:45]:
Josh, in your world, how do you define security, the scope of security for that matter? Because you’ve got various entities connected in the ecosystem. It could be the management, the faculty, staff, students, even their parents and/or anyone who’s near and dear to them who are concerned about security. And if there is at all a breach, then you feel that you are in the spotlight. So what do you do to draw a circle around all possible things related to security and work within that cocoon, if you will?

A decentralized campus and personal responsibility 8:17

Joshua Beeman [00:08:17]:
I think every institution is probably different. At Penn, we’re extremely decentralized. So we’ve got 12 different schools. You might be familiar with the Wharton Business School, the Annenberg School of Communications, etc. And for us, each one of those 12 schools is independently responsible for ensuring that its unique IT needs are met in a way that suits the school the best. So the result is that each school could vary in their expectations from faculty or from staff, or from students or their parents about the degree of security literacy or competency that they require. But what I think is true for all 12 schools, and probably regardless of institution, is the realization that we’re increasingly in an age where personal responsibility often trumps some kind of centralized control. And what I mean by that is that a user could invoke a high-performance cluster at Amazon with a couple of clicks of a mouse in a second, or they could download huge chunks of data onto a personal device, a small personal device that’s easily lost, just by signing up for some free service. And so the user can do all this without the help or knowledge, frankly, of the local IT department. In situations like this, the greatest risk we’re facing is ourselves. I think that given all those risks, the ones that David’s already talked about, BYOD or bring your own device, self-service cloud computing, our goal is we’ve gotta be able to provide secure, reasonable alternatives. We’ve gotta have policy and guidance that is clear and concise so people actually understand what we want them to do. And we’ve got to have strategic layered security controls that don’t counteract our mission, which is, for Penn, eminence in teaching, research, and service.

Sanjog Aul [00:10:20]:
Dave, where do you think the actual security breaches originate from? Is it truly the mindset of those individuals who are expected to observe whatever has been suggested as a set of guidelines what to do or not to do? Or can it primarily be attributed to the effort, the resources, the planning that we do around security management in an education setting?

David Sherry [00:10:47]:
That’s a one hour question itself, Sanjog Aul. But I’m going to echo what Josh said, is that every school, while at the end of the day they are for teaching and research and learning, every school is a little bit different. And the bigger schools, the public schools versus private, community college versus research institutions, all have it a little bit different. So I can speak from my role here at Brown University. We have security events across the board. It’s because of someone answering a phishing email even though we work very hard at our awareness efforts. Some people just fall for it, a well-crafted one. Certainly, we have compromises that occur because people don’t follow a standard build or they throw up a server that has outdated protection without notifying the network group. And we work very hard at trying to get the word out about centralizing things is not a bad thing. We want to be a compelling option to them. We don’t tell them that they have to come under the centralized management, but we want to certainly support them in their efforts to do what they need to do to get the teaching and the research and the learning done. And my tactic here is always when any group comes whether it’s students, faculty, staff, we’ve worked with alumni, we’ve worked with donors, if they come with us with a solution that they’re looking for or an idea that they have, my first blush is never to say no. I always say, yeah, we can get that done, but it’s going to cost you some money, cost you some time, cost you some resources, and then leave the decision up to them as to whether they want to pursue it from a secure manner. Well, that’s when we have to step in and say, no, that can’t be done because it brings too much risk to the university.

Sanjog Aul [00:12:26]:
Josh, are there any weakest links if you will which you can outright identify whether it is your organization, institution, or across the board?

Joshua Beeman [00:12:38]:
No, I think I would just return to the comment that I made before about being in an age of personal responsibility. I really think that IT organizations have to be more diligent than ever in educating users about the power that they now have to bypass any security control that we could implement.

Sanjog Aul [00:13:04]:
Dave, how much immunity does a CISO have in an education setting where we know that there are constituents who are impacting security and overall vulnerability of the organization?

David Sherry [00:13:16]:
I don’t know if we can ever have 100% immunity, but it’s our responsibility to make sure that we’re identifying the risks, identifying that the university is never going to be 100% secure in a decentralized environment and, as going with what Josh says, an era of personal responsibility. So we need to maintain to the executives at the university all the methods and processes and governance that we’re putting in to keep the university at as low a risk tolerance as we possibly can and as secure as we possibly can, and preparing them that something can go wrong. And on top of that, having obviously a very robust, scalable, well-practiced incident response process, so that when something does happen, we can shut it down quickly and show our worth in quickly mitigating the danger that has come in.

Sanjog Aul [00:16:14]:
So, Josh, let’s compare business setting and educational institution setting, and let’s see where would you do things which are incrementally more complex and/or more stringent and/or more flexible for that matter in order to meet the security needs for that organization that you run.

Why mission shapes security policy 16:38

Joshua Beeman [00:16:38]:
Well, Sanjog Aul, I think it’s a cliche, but policy is really the foundation of information security. And good policy, the rules that tell us what we can and can’t do in our organization, to be effective derive directly from our mission. So I think if you want to talk about the differences between business and higher ed, what you have to do is look at the differences in their mission. As I think both David and I alluded to earlier, the missions of our organization relate to teaching, to research, and to service. And those are not specifically the goals of business. So the technologies themselves that we have at our disposal are the same. But I think that the ways that we implement them and have to think about implementing them are different. So I like to joke that when it comes to BYOD, this is like one space where higher education has been in the forefront for decades. Because every fall for the last 20+ years, thousands of new students have been showing up with personal computers and smartphones and gaming devices and who knows what else, expecting a secure, reliable network, expecting technical support for what is undoubtedly a very diverse set of operating systems and applications. So I think to that extent, while the technologies are the same, the rules that we put in place have to reflect the realities of our missions. And to that degree, I think, in some ways education, it comes back to education and influence rather than explicit control in a place where our primary mission is about enhancing collaboration and openness between organizations.

Sanjog Aul [00:18:39]:
So would you say, David, this is for you. Do you think that we will have to balance between risks and benefits based on what Josh just mentioned that because we have to impart education and because we have to have a collaborative environment, we could potentially not exactly ignore, but live with a certain amount of risk in the interest of what this collaboration would bring to us?

What people can, may, and must do 19:06

David Sherry [00:19:06]:
Absolutely. Every decision we make is looking at the risks and the benefits and trying to balance that. Certainly, some things are high risk and we have to take away some of the access to it. If we have HIPAA data or we have research protection data, human subject data that is extremely high risk and we have to put high security controls on it. But if we’re talking about the residential network where the students come in, as Josh said, with multiple devices and Apple TV and gaming stations, wireless routers, we let them; the risk is a little bit lower there because they’re separated from our network and we can allow them to do some more things. And going along with what Josh said about having good policy, we look at it here that there’s kind of three concentric circles. The biggest one is what people can do. The smaller one inside is what people may do. And then in the middle part is the smallest one is what people must do. And the difference between that must circle and the may circle, that’s where the policy comes into place. Must is a mandate. We must protect HIPAA data. There’s just no questions about that. It needs to be centralized and it needs to be mandated and assessed periodically and regularly. It’s the things that people may do, but there may be a risk to it or some kind of regulation that we need to follow, and that’s where the policy comes into place. We seem to focus on that more than anything else. We think we have a good handle on the must, but it’s the may. I came out of financial services before I was in higher ed. And when I was in financial services as the security officer, I had the ability to say, thou shalt do this or thou shalt not do that. And I come into higher ed and it’s like jumping into the wild, wild west, which makes my job so fulfilling and so much fun. But it’s also more challenging trying to get the balance of where the risk is to the university, where the risk is to the certain department, faculty member, or student, and making sure at the end of the day that the university is the one that is accepting the risk tolerance and protecting what we need to.

Joshua Beeman [00:21:00]:
I think that as we make these calculations, we can do an equivalent job to balance risk. We just have to think of our constituents and figure out ways to provide those security controls that are meaningful to our populations. So, if we’re talking about free services, or BYOD, we’ve got to figure out a way to provide reasonable alternatives or secure alternatives, and to inform people about the policies that we have. I truly believe that people want to do the right thing on our campus. We’ve just got to enable them to do so.

Sanjog Aul [00:21:42]:
Now, Josh, the fact we spoke about setting the policies. So do you think if you set the right policy, is that enough? Or if you get it adopted by whatever means that you get it adopted, is that where you are measured for, in terms of your success in making sure that you created a secure environment? And if you do try to go and implement those policies or get that policy adopted, is that done in the same authoritative approach or saying that if you don’t do it, then x, y, z would happen, so it’s almost like a threat. Do you think the community, the student community specifically respond very well to that?

Education over threats 22:21

Joshua Beeman [00:22:21]:
No. I don’t think people respond to threats. The problems that we typically see stem from a lack of education or a lack of understanding about the technologies or university policies. Threats are not going to help a student system. They’re not going to help solve a problem with a student system that’s been compromised because of some application or operating system vulnerability. Education is going to help that. Education coupled with effective detection, with automated security controls, these things are far more practical and effective than some kind of authoritarian sort of threat about what happens if you don’t do something.

Sanjog Aul [00:23:05]:
So, David, I’m a parent, and I’m sure many of the listeners are too. I’m not truly concerned that if my son’s or daughter’s computer gets hacked. I’m more concerned about the abuse that could happen online and using online resources or on campus. With that said, the security policy itself may not help some sort of sniffing or some sort of intervention in terms of if the semantics of what is being exchanged in terms of collaboration is not looked at closely, then a lot of things people could get away with doing something which is wrong and which in a way is breaching the security that you were supposed to safeguard. What do you do about that?

David Sherry [00:23:49]:
We certainly can’t be sniffing or doing log management on everything that’s being shared throughout a university that is quickly ramping up, getting bigger with online courses and global presence. A lot of it still goes back to the awareness and the policies that we have to put out and raise the awareness level amongst our constituents. I do speak with a lot of parents. One of the times of the year that I really enjoy is the beginning of the new academic year where the freshmen show up with their parents, and I listen to the parents’ concerns about online safety and what resources they can get access to, what they shouldn’t, how can we respond to them quickly. And we do talk about copyright and plagiarism because in an area of right clicking and copying, the academic integrity of the institution can be at harm when students fall prey to something like that. Going back a little bit to how we get this out, two of the key things I think that have been successful in any security stuff that I’ve been on is persuasion and influence, more than threats. It’s telling them it’s the right thing for the institution or the right thing for the enterprise to do in such and such a way and that I’m here to help them. I tell them I’m not here to disable or stop any university business. I’m here to securely enable it to make sure that we’re going to go forward. Brown is going to be celebrating its 250th anniversary in 2 years, and I tell them that I want it to go on for another 250 years. Certainly, having a secure environment, doing the right thing, and making sure that we don’t have any compromises that are going to stop kids and their parents from applying here is one of the key things that I have to be focusing on.

Sanjog Aul [00:25:31]:
Josh, you spoke about policy making and, of course, implementing it. Where does this policy originate from? Do we actually listen to the students and/or parents and other constituents and then make this policy, or is it more of an ivory-tower, top-down approach to what should be working if we were to make it secure? Because even that definition of what is secure may be morphing under your feet as you go about implementing and even envisioning such policies.

Joshua Beeman [00:25:59]:
Policy, something that may be dry to many people, is something that’s near and dear to my heart. It’s such a valuable tool for us in getting our jobs done. It is, I believe, one of the most collaborative processes that I work in, and it takes input from everywhere. We take input from peers. We take input from IT people, from non-IT people, from users, from family. We’ve got to respect, as David alluded to earlier in talking about the musts, there are certain things that are simply not optional, right? So we’ve got to pay attention to regulatory spaces and to compliance requirements. But we’ve also got requirements that are unique to an individual organization or requirements that are unique to the character of our institution, to an institution like Penn. And we’ve got to reflect all of that in policy. I heard a Supreme Court judge talking about the law once, and he said that the most dangerous kind of law was one that people willfully did not follow. And I really think about that a lot when I’m crafting policy. We try, and again, I think David alluded to this as well, to create policy that is reasonable and meaningful for people while setting out the expectations that are necessary with regards to security and securing the data, and the digital assets of our university.

Sanjog Aul [00:27:32]:
So Josh, suppose you built a policy and while you made your best effort, a few components of the same don’t fly and they are not getting adopted. Once you have already implemented it, it would be kind of interesting for you to revoke it or revoke pieces of it because it’s not working. How do you manage this expectation setting? And in case you have to revoke certain things, how do you reset those expectations?

Joshua Beeman [00:27:58]:
It should be an iterative process. Everyone’s policy process is different. I can talk about ours at Penn. One of the ways that we try and avoid having to revoke something, and that doesn’t mean that such things don’t happen, but it can be confusing to organizations and to the constituents when that happens. Policy isn’t the most exciting thing for people, so you really do want to try and get it clear the first time out. One of the ways that we combat that is we have a public review period where we actually shop draft policy around in advance. So we give people a chance to react to it. And the other way that you hopefully sidestep that issue is you have enough people at the table the first round in that you’re really getting a diverse set of viewpoints. But I will say, the technology changes. We’ve seen policy that sets out requirements for technological things that don’t exist anymore. And you’ve got to update that policy. It’s got to be a living, breathing thing or it stops being meaningful to people.

Sanjog Aul [00:29:03]:
So, David, when we build our course curriculum and different programs, we keep the customer in mind, which is the student and what the world looks like outside in terms of what’s going to be more valuable for an individual to put on their resume. But when we are talking about security, do we really keep that customer in mind in terms of what they want? Because typically, the adoption happens bottom-up.

Sanjog Aul [00:31:01]:
The policies are being made for other things or business models are made to serve the customer. But when it comes to security, how much of that is ivory-tower versus how much of the time did you actually include the customer, which is the student and/or perhaps their parents, to make this happen because adoption happens bottom-up?

Involving students in policy 31:23

David Sherry [00:31:23]:
Well, ivory-tower does not work, Sanjog Aul. That was proven long before I got here and I quickly learned that mandates from the top usually just either get comments of derision or just flat out denied or not even followed. So our process follows very similar to what Josh described at Penn that we talk to all the constituents, the alums, the corporation members, donors, students have a big role in any policy that we put out. The faculty have a big role in it as well. We do requests for comments. We do town hall sessions. We put it up on our website for public review. I was told when I got here to get a policy approved, it takes one year per page. And that has pretty much held true in what I’ve seen over the past 4 years. We certainly do keep the customer in mind. I take what the students say very, very seriously. We have a campus paper that is not shy in voicing their concerns or their complaints about anything that goes on on campus, especially IT and security, so I certainly meet with student groups and focus groups to find out where the balance is. They understand that, when they’re on campus and maybe in the academic network and in the research network, it gets a little bit more corporate than what they might want, but they have to understand that. But when they’re in the campus network, the residential network, they don’t expect a corporate environment nor should they. And we find the balance of what they need, what our firewall rules will be, maybe what the bandwidth, maybe the protocols that we have to shut off and open up to make them happy. And, certainly, we do keep them in mind because they’re the loudest voice. They’re the ones paying the tuition and keeping the university going.

Sanjog Aul [00:33:02]:
So, Josh, the question for you is that while we all want to grow as a business and so does an educational institution, and then we want to innovate in that area. While we spoke about innovation from a technology standpoint, where you’re trying to bring BYOD and make it easier and that’s posing challenges, but besides that, what other innovations are actually which may look good in terms of bringing growth to the organization, but also introducing risk? And what are you trying to do to manage the balance between growth and risk?

Innovation, social media, and risk 33:34

Joshua Beeman [00:33:34]:
I think this is just a great question, because it really talks to some of the challenges that we’re facing every day. The landscape is changing very, very quickly. So you mentioned BYOD. Online learning just continues to expand and evolve in our space, with profound impacts, cloud-based computing, and social media. To talk about social media, for example, I have this photograph that I talk about all the time, in presentations. It was a photo that I saw, from a visit, of Katy Perry visiting Mark Zuckerberg’s office at Facebook. And it was a shot of the inside of Mark Zuckerberg’s office. And above the door, exiting the office, he had two signs. One said, “Move fast and break things.” And the other said, “Done is better than perfect.” I love this. I think that these are the two messages that Zuckerberg wanted Facebook employees to take with them as they’re leaving his office. And it’s a great example of how technologies that we embrace, in the case of social media, have amazing potential and capabilities to improve how we’re fulfilling our mission. They also can have goals that are possibly totally misaligned with our own. Facebook’s mission is to connect people globally, but it’s also to be the first to market with new ideas. And those values, at times, are in conflict with my university’s values of privacy and security. So we’ve got to balance this potential with the risk. I think that means being aware of these really innovative and transformative technologies, and not being afraid to embrace them, but doing so deliberately, and hopefully knowledgeably. So I’ll give you just a brief example at Penn: for social media that resulted in us publishing guidance on how to use social media safely at work for faculty and staff. We also created a series of tips for how to create a secure Facebook profile for everybody, for students, faculty, and staff. And then of course we promote the heck out of these tools and hopefully we’re better securing an experience that people are excited about but not inhibiting or diluting the power of these technologies and their innovations.

Sanjog Aul [00:36:10]:
So, David, we would love to get your input on this too. So do you think that there is something on the horizon that you see where whatever we are dealing with now is going to get compounded?

David Sherry [00:36:25]:
Yes, that’s a good question. I think the thing that we’re dealing with now that has changed dramatically just in my short time here is what is the primary network? When I was here with wired connections and there was wireless access in certain locations, it’s getting to the point where the wireless network is the primary network and it’s expected to be ubiquitous over the entire campus and even some of the surrounding areas where Brown has a footprint or an influence, and that certainly brings a lot greater challenges with compromises, man-in-the-middle attacks, use of it by non-Brown people. That’s certainly changing the balance of where we’re going as we get less wired and more wireless. And I agree with Josh that this is a challenge every day. I’m not going to beat the BYOD, we’ve talked about that. Cloud with Amazon services being able to spin up so quickly. Sometimes they don’t want to wait for the security or the central IT team to come up with a solution. Certainly “done is better than perfect” is a great mantra for Facebook, but in a security world, we have to get pretty close to perfect when we roll something out, knowing what the ramifications are if it’s not correct. And I’ve always told my staff and talk about it in public presentations that I think the security team’s reputation is a game of chutes and ladders. We climb our way up the ladders as quick as we can when we have a good project. But if we have one minor slip up, we go down that chute and we’re starting all over again right from the bottom. So, reputation management of the security team is paramount when we’re making the balances of growth and risk too. Do we speed a project up to get it in a little quicker and maybe have some issues that we have to hit downstream or do we say, hey, we need an extra month to get this in correctly knowing that the university is going to benefit from that.

Sanjog Aul [00:38:19]:
Now Josh, do you think what we are doing today, for the most part what we spoke about, somehow seems like it is confined to the campus environment? But we truly have a huge percentage of people who, while they may be joining an organization or an educational institution for getting their degrees, the majority of the actual effort, collaboration, and you name it, might be happening from places outside of the campus. To what degree do you think a CISO of an educational institution is responsible, and will be able to take control of the security of that customer who’s a student even though that person is not physically present in front of their eyes? And are you willing to take the responsibility for that?

Joshua Beeman [00:39:04]:
Well, I think that the scope of at least my responsibility is reasonably clear, and that is to protect Penn’s digital assets. So the location of the users really is not the most important thing. If institutions of higher learning continue to evolve and move more and more content and resources online, where the brick-and-mortar school is ancillary or complementary to that, that’s just the way it is. In the same way that I don’t have the luxury of only worrying about attackers, or people who violate policy that are physically located on campus, I have to also worry about protecting anyone who’s accessing any of our data anywhere.

Sanjog Aul [00:39:57]:
Now, David, you have a team, I’m assuming, which takes care of this number of security issues and safeguards security for the educational institution and its students. What’s your mandate for your team when a person from your team could get fired for not listening or adhering to some of the mandates you have? Let’s inventory those.

The security team’s mission 40:20

David Sherry [00:40:20]:
My team is very similar to Josh’s. Our mission is that we’re here to protect the university and the university’s digital assets and even all information assets, hard copy and digital. My team has a threefold mission. And that’s one, to build a secure and robust security architecture, which encompasses across the entire university, both centralized and decentralized. We also have a mission of being the experts in proactive subject matter expertise in all matters of security. So it’s a forward-facing team. It’s the architects who may be the plumbing and the people things behind the scenes, but we also have a forward-facing mission that we need to be the people to go to for lifecycle security projects and security expertise on upcoming projects. And then we also have the mission of creating a culture of security awareness. We take that mission extremely seriously. I used to laugh and tell people that my role in many ways is protecting the university from the world, but at the same time protecting the world from the university. If I see IT staff, security staff, or anyone overstepping their bounds, not following our acceptable use policy or our student conduct or employee conduct policies and doing something wrong, we have to make an example of ourselves to show that we take our role seriously, and we’re not going to just look the other way if someone oversteps their bounds and may do something outside of our security parameters.

Sanjog Aul [00:41:47]:
Josh, typically a good leader and/or for that matter a great leader has a vision which is used to make people do things versus threatening them that if you don’t do something bad will happen to you. So what is the kind of vision that you are painting for not only your staff but for the organization as a whole so that the people follow? And is there a vision? Is there a possibility for you to paint a vision given that the things are moving at such warp speed?

Joshua Beeman [00:42:19]:
Well, I certainly do think it’s possible to paint a vision even though we now live in a world where technological advances are happening very, very quickly. I think leadership is a very interesting thing. People come to it from different places. One thing that I observe works is that the kinds of people that I work with in information security are really drawn to and motivated by some of the rapid technological change and the challenges that we’re talking about. These are people who are very excited about the future, and they’re very excited about having a positive impact on the future. In terms of the broader picture of how one motivates and leads people to a security vision, I think that it’s really very much a question of understanding what motivates people and having a good handle around the facts of your environment. I’ve been very fortunate to have success at Penn simply by talking to the various constituents in the schools and cost centers at Penn about what their needs are, but also to understand the core mission of the university and then align the interests of people with the long-term goals. David alluded to this earlier, but I cannot underemphasize the importance of conversations with peers, conversations with subject matter experts about where the world is going and where security is going. It is a rapidly developing field, and when one identifies the proper direction, then one needs to build the proper enthusiasm to get there.

Sanjog Aul [00:44:17]:
David, are there any concerns that have been explicitly expressed by the students and/or parents or someone else who’s also either a customer or an influencer where you feel you don’t have a solution to this today?

David Sherry [00:44:32]:
I think that happens a lot, Sanjog Aul, but mainly it’s because a student or a parent will come out of nowhere with a question that we haven’t thought about.

Sanjog Aul [00:44:42]:
What would be some of those things that you may be struggling with? I’m not saying that you have to literally open up the vulnerabilities, but what are some of the things that people ask of you which you feel realistically cannot be delivered today or the technology and/or the way the educational system works you simply are not in a position to deliver?

David Sherry [00:45:04]:
Well, some parents want access to systems or access to student data that is covered under FERPA. By regulatory mandates, we obviously cannot provide that without stipulations from the students. We have a very enthusiastic and talented computer science community here and sometimes the computer science students are always testing our limits and using anonymizing proxies and Tor networks and possibly opening up ports and protocols that we feel are not necessary. So sometimes we have to push back on that and circle around, talk to them, see what their needs are and come up with a more secure solution.

Sanjog Aul [00:48:12]:
So when it comes to the budgets, the budgets that we talk about in terms of the allocation of budgets with respect to security, we always see that as a stepchild treatment when it comes to what dollars you need, what resources you need in order to make things happen. What do you think is the situation? Is that hearsay, or would you say something which there’s some reality to it, Josh?

Joshua Beeman [00:48:46]:
I don’t know anyone in the field that would ever say no to more funding. At least at Penn, we’re a not-for-profit institution. It’s a challenge always to balance limited resources. As we’ve been talking about for the last 45 minutes or so, we live in a world where attacks are constantly evolving. I think what’s clear to me is that Penn does view security as essential and is totally committed to protecting the information of both the individuals and the constituents, but also the university at large. From my experience, that is the case at my peer institutions as well; they understand the value of this. The devil’s in the details, how much money people are getting. I’m sure people would always want more.

Sanjog Aul [00:49:43]:
So, David, this is a question for you. Can we make this statement confidently that while people want the world out of your group, which is to keep the educational institution and the students secure, you would say that you got everything you need? Or if I gave you this forum to appeal to your management, to say, give me $1,000,000 more, I can show you what I can do? Would your answer be yes?

Justifying security spend without fear 50:15

David Sherry [00:50:15]:
Yes. We’re constantly doing that. I’m going to echo once again what Josh said that everything is constantly changing. The risks are changing and the risk profile is changing. So all budgets are tight. We live in a national and global economy where people are asking us to do more with less. So we need to substantiate what our needs are and show institutional value by writing strong business plans, that you can’t use fear, uncertainty, and doubt. Here at Brown is part of the overall risk management framework, so I do have a voice and I do have an audience where I can put forward what I need, but it needs to be documented and substantiated for what the value is going to be. Sometimes it’s hard to get an ROI on a security project; if you spend a lot of money and resources and time and maybe FTE and professional services and then nothing goes wrong afterwards, you have to balance; you have to teach them that nothing went wrong because of what we did and they don’t think that we’ve wasted the university’s resources by spending a hefty amount of money and then not having anything occur after it. So I think, as I stated, having almost MBA-like thinking, have strong business plans showing the value, and having a 3-5 year vision of the overall security architecture knowing that there could be changes as something occurs, is paramount.

Sanjog Aul [00:51:34]:
Josh, we have seen that organizations put very strong policies and even there are people who are buying in, but the very business process by which something of value is created, whether it’s education being imparted or other back office information systems or other work-related items are conducted, it introduces vulnerability. Do you think as part of what you do, you also have influence and/or control over how business processes run and you can point fingers and other people will listen and actually spend the money, time, and effort to fix it in order for you to meet your goals? Or do you find challenge there?

Measuring risk with data not opinion 52:12

Joshua Beeman [00:52:12]:
I would say that this is one area where I really and truly believe things have gotten better, and in no small part due to the increasing publicity and general public awareness about computer security. At Penn, we have a voluntary process that every school and cost center participates in. It’s called SPIA, which stands for Security and Privacy Impact Assessment. The goal is to prompt organizations on an annual basis to identify and improve the security around applications and databases which house any kind of sensitive data. The first few years of this really involved more coaxing to get participation. But now everyone is involved, and I can say confidently that we’ve got every single school and center now at Penn involved in this project. Everyone involved understands that this is a tool that not only helps management gain awareness, but by participating in this it drives really important improvements in their organization and ultimately lowers their risk. David used this great term, fear, uncertainty, doubt or FUD. Without data, it’s just another opinion. If you’ve got strong data, if you can help people see and identify where the issues are, then it’s much easier to align resources towards them.

Sanjog Aul [00:53:40]:
David, how much security is enough in your educational institution type of setting? What are the standards and/or benchmarks that you look at as a holy grail and work towards that? Or is there even one which you think, once achieved, would you call it done and take a 2-week vacation?

David Sherry [00:54:00]:
A 2-week vacation would be really nice. Well, how much security is enough? I use a cross-university model to figure that out with strong relationships with our internal audit team and external and our legal group and also our chief risk officer. We as a group come up with how much security is enough and we align that with the federal and state regulations that we must comply with and we compare ourselves to different frameworks. I like to lean towards the ISO 27000 series. I think it covers everything that we need. We look a lot at NIST because we have a lot of federal grants and federal money that comes in that we have to respond to. We look at things holistically throughout all the models. We try to build extremely strong processes. We have a big strong ITIL working relationship here that we’re trying to go through. The benchmarks and the standards need to meet the needs of the university, not just from the security perspective, but also from the audit, legal, and risk perspectives as well.

Sanjog Aul [00:55:05]:
Josh, for you, what is that you will do in terms of leadership and executive sponsorship and executional excellence in order to build a strong foundation for imparting education to everyone starting from say higher ed and/or elementary students?

Joshua Beeman [00:55:28]:
I don’t think that there’s anything unique about the information security sphere in this regard. If I understand the question to be, successful information security people have to be able to explain security issues in a way that is meaningful to all levels of their organization, obviously including leadership. And that means having the right tools, the right processes and the right people to be able to effectively describe our risks and to be able to measure our risks. David just alluded to a process that he has, and I alluded earlier to a process that I have. When we can do that, the right decisions make themselves known. At Penn, I’m fortunate that I can point to an example of a key process like this. It’s probably because we’re so decentralized. We’ve got a really strong IT governance structure that includes information security governance and it encourages this dialogue and debate across all the university’s stakeholders. Anything that encourages communication, discussion, and prioritization by senior management is ultimately going to facilitate a strong and secure organization.

Contributors

David Sherry

David Sherry, CISO, Brown University

David Sherry is the Chief Information Security Officer at Brown University, with university-wide responsibility and authority regarding matters of information security and privacy. He leads the Information Security Group, charged with the d... More   View all posts
Joshua Beeman

Joshua Beeman, CISO, University of Pennsylvania

Joshua Beeman is the University of Pennsylvania's Information Security Officer. The Office of Information Security at Penn establishes, implements and maintains security programs to assist with the protection of computing resources and asso... More   View all posts
Add Comment
Click here to post a comment

Advertisement

TECHM - DNA - MPU01 - 300x250
David Sherry