The future in which artificial intelligence shapes America's city is here. New York City's Chief Technology Officer, Matthew Fraser, joins American Banker's Penny Crosman to discuss how NYC is on the forefront of the AI revolution, harnessing the technology to protect public safety, create cyber resilience in both the public and private sectors, and guard against the next health outbreak.
Transcription:
Penny Crosman (00:10):
Welcome. I'm Penny Crosman with American Banker. How does New York City's tech chief see technology shaping this great city of ours? I'm here today with Matt Fraser, Chief Technology Officer of New York City. Welcome, Matt.
Matt Fraser (00:24):
A pleasure to be here.
Penny Crosman (00:25):
Thank you. And we're going to talk about how he uses technology in his organization, how he thinks about I, how he is defending our city from cybersecurity attacks, and how he partners with other businesses in the city to work together to defend all of us from cybersecurity threats. So thank you for joining us today. So I guess to start with, during the pandemic, we had this very unusual challenge of access to the internet and millions of people had to leave their offices, leave their schools, go back to their apartments and try to use wifi from there. As head of technology for New York City. What were some of the challenges during that time and what did you do to address them?
Matt Fraser (01:16):
Yeah, so we can speak from the 2022 perspective going forward because that's when the administration took seat. When we came in, there was a program called the internet master plan, and the internet master plan was focused on providing broadband in areas where they typically have been disconnected, areas where the average income met low income thresholds. Now, the internet master plan was projected to be a $2 billion program. The first phase of the program was funded for $157 million and it was launched right before the actual pandemic got wings and became a big problem. And throughout the pandemic, the program didn't deliver one single connection to a household. So when we stepped in January of 2022, one of the things that we took stock of is like, let's take a look at the digital landscape. Let's look at digital equity and let's look at the communities that historically have been left behind and see how fair that connectivity looks.
(02:18)
And what we used as a test case was we took public housing as an example. About 40% to 50% of those that lived in public housing didn't have access to high-speed broadband. Now you take that and you put that in concert with the fact that you have school-aged children that live in some of these households. They were sent home with devices. And when you walk into some of these buildings, they're pre-war buildings where you can't get connectivity in from a cellular perspective. So now they're at home, they're disconnected, they can't maintain their education, and also from a social perspective, they can't communicate with their peers because everyone else is at home. We found that to be a big challenge. So what we did is we looked at, alright, instead of talking about just what the problem was, we tried to figure out how we can solve the problem.
(03:06)
So again, using public housing as a test case, we looked at bringing in our big cable franchises. So you think about folks like Charter Communication, Spectrum, and then we also look at bringing in Optimum and Verizon. We said, listen, let's figure out how we solve the challenge. You have infrastructure, we have people that need access to connectivity. How can we work together to do that? So within one year of being in office, we launched a program called Big Apple Connect. And what Big Apple Connect did was it went across the entire universe of those that lived in public housing and it gave them options so they can have free access to high-speed broadband and basic tv. So now they didn't have to choose between putting food in the fridge or paying their cable bill. That cable was now provided to them at no cost. And for us it was a critical necessity to ensure that they had it because it gave the kids that lived in those areas, the older adults that needed the access to telehealth or regular adults that were looking to apply for jobs, a connectivity and pathway out, not enough, but it's a good place to start and we were able to do that within one year.
Penny Crosman (04:12):
That's great. Did you end up partnering with Verizon or Spectrum or Optimum?
Matt Fraser (04:17):
So the way that we did it is we took a look at the rates, the average market rates of what they were charging, and we looked at the Federal Affordable Connectivity program as a comparable, and the federal government right now has a program for 29.99. It reimburses Lifeline Services, so you can choose to get a reimbursement for your house, internet connection or your cell phone connection. So what we did was we brought the cable carriers to the table and we said, we need to provide a program to public housing to give high-speed broadband, and we need the model to be based price-wise off what's available at a federal level. So we were able to partner with all three of our major cable franchises in the city, and we struck a deal where in public housing for the same rate that the federal government reimburses, they got not only just access to free broadband, but they also got access to free basic TV as well.
Penny Crosman (05:14):
That's pretty amazing. So I want to ask you about AI, because AI is the big thing everybody is interested in thinking about. I guess from a personal perspective, some people think of AI as something scary that the machines are getting more sentient, more powerful, something to be concerned about. Other people see limitless opportunity for using AI to make jobs more intelligent, create automated processes, make life better for people. What's your personal attitude toward ai?
Matt Fraser (05:55):
So I think a lot of the stigmatism behind AI comes, I'm a kid of the 1980s, so being a kid of that generation, how could you not know a movie like the Terminator? So when people hear artificial intelligence, they think about Skynet, CyberDyne Systems and them creating autonomous robots to walk around and take place of humanity. I think that we're a long way from technology being at that level, but artificial intelligence as a whole gives us the capability to process work faster and in some cases remove human involvement in areas that don't require a human decision. When you think about processing paperwork, being able to have a system that can go through and validate that all of the required information has been provided and that it's in the format that's necessary and it's actually ready for human review, and then being able to preload that with a set of predeterminations that says of the information that's provided, it seems like this application is 70% more likely to be approved.
(07:08)
You see, when we look at artificial intelligence, it's existed for a long time and some of the places where you can see that, especially in the financial space is when someone applies for a credit card, there's a lot of determinations that are made around approvals, denials, and then whether human interaction is required to review the file or not. Very similarly in the cybersecurity space, artificial intelligence has been employed for a number of years to help secure everything from computers to cell phones, making decisions in real time about what sort of behavior should it allow or deny. So artificial intelligence, at least the way that most people look at it today is they look at these large language models like ChatGPT, and I think that's the thing that people get really concerned about. It's like these models provide information and it almost seems like the information that comes out comes from an authoritative source, but you don't know where the information is sourced from, you don't know how valid the information is. And what folks are very concerned about is that by leveraging models like that, you can spread disinformation very quickly and you can form an opinion about something that's not true because people depend on the technology more than they depend on proving that the information is right. So I think for us, artificial intelligence will prove to be a useful tool, just like all technology that's come out to this point. And it's on us to make sure that we have guardrails up. So when we deploy it, we deploy it in ethical and more ways.
Penny Crosman (08:37):
So today it's not in place, but it's something you're thinking about using?
Matt Fraser (08:42):
No. So today it's in place, and I can give you some practical examples, right? When you look at New York City from a cyber perspective, we have over a million endpoints, and our weekly average in cyber related events is about 80 billion. So if you think about how many human hours would be required to review all 80 billion events, that's a significant amount of human hours. Needless to say from a cost perspective, it doesn't make sense to staff at those levels, nor does it make sense to have humans review each one of those events to make sure that they're valid or not. We use artificial intelligence in that way to look at each event, assign a risk score, look at the behavior, seeing if we can auto remediate what's happening or stop the malicious behavior or allow behavior that looks malicious, but it's actually good. And then if it's significant enough, raise a flag so that a person can look at it. So out of the 80 billion events that we see weekly or weekly average, we distill that down to less than 50 things that someone has to look at at any given week. So it's a good example of how artificial intelligence is being used to help us work smarter and distilling the things that are important from the noise.
Penny Crosman (09:56):
Can you say anything about the types of threats that New York City has been seeing lately?
Matt Fraser (10:03):
What we see as the city is the same thing that the rest of the global community sees. Escalated activity around distributed denial of service attacks, phishing and business email compromise is what we tend to see. In addition to that, you have your periodic brute force attacks where someone is just trying to penetrate through and coming into the network. On any given day, there isn't a breadbox size of the type of attack that we see. It ebbs and flows like a tide every day. It's a little bit different.
Penny Crosman (10:40):
I was thinking last week was the week of September 11th, and I'm sure everyone was on heightened alert. I mean, we were as commuters. Do you see more threats that week typically?
Matt Fraser (10:52):
So I'd say we exercise persistent vigilance. So that day versus any other day, it's hard to say specifically how the threat ebbs and flows, but we are particularly diligent around those notable days because that's when bad actors look to create chaos, right? It's like opening an old wound. I think for us, making sure that we are, regardless of the day or the time of year, we are persistently looking at the threat landscape and making sure that we're blocking as much as we can and responding as quickly as we can consistently, I think that's how we keep New York City safe.
Penny Crosman (11:35):
It seems like often cities, especially small cities, are targets for cyber hackers because they often don't have state-of-the-art IT equipment. They often have older equipment and firewalls that maybe aren't that strong. Do you feel it's imperative for New York to kind of keep the infrastructure current with state-of-the-art best firewalls and the best defense shields possible?
Matt Fraser (12:05):
Yeah. So it's not always about the technology. Sometimes it's about poor hygiene. So you could have the latest and greatest technology, but if you don't patch it frequently, you'll find yourself in a position where that technology, although it's new and great, a day, two days, a week later, it'll be just as susceptible to compromise as an older system. So what we focus on is proper hygiene. It's like when you're a kid, you get taught to brush your teeth, take your shower, do all the things that are necessary to maintain your health. We do those things from a system perspective. We make sure that the technology that we have deployed continuously is patched and updated again. And then we also make sure that when the technology ages beyond support, we either have a compensating control to make sure that it's safe or we get rid of it and replace it with something that's new.
(12:54)
Part of the complexity for smaller cities is that when it comes to financial resources and technical talent, it may be a challenge for them to get resources at scale to defend that landscape, which is one of the reasons why at the start of the [Adams] administration, we launched the joint Security Operations center in collaboration with the state and the governor. And part of the reason was for New York State, New York City Cyber Command is an entity that existed for the last better part of a decade, whereas across the broader state, a similar entity did not exist. So as we looked at building capability at a state level, we said, why not just borrow from the construct that exists in New York City? And in fact it might be better if we do it as a collaboration so that the smaller cities across the state don't have to build their own shops, they can join one central shop.
(13:51)
And what you get out of that is you get a lot of economies of scale. So if you're looking to invest in things like security technologies for your desktop, you're looking to invest in security operation center technology instead of each small city having to make that investment themselves, they can join a collective and they can value and benefit from the economy of scale. Another big challenge is when funds come down at a federal level, the state has to make a determination on how it distributes those funds across those funds across the cities, across the state. And for us, if we had a centralized construct, instead of making determinations around one small county or one small city in the state requires 10 million to make an investment into something that it hasn't invested in the last 20 years, you invest in a central place so the entire state can benefit from that investment.
Penny Crosman (14:44):
That makes a lot of sense. Now, I've heard you also collaborate with businesses in New York, like large banks. Can you tell me a little bit about that on the cybersecurity front?
Matt Fraser (14:53):
That is very true. So on both cybersecurity and physical security, and when you look at security, it's an exercise of community. So we can be as strong as we want as a city, but 90% of our constituency is the people of New York. So you think about everything from your banks to your supermarkets, to your department stores, those are all privately run entities. Starbucks may not be considered a critical business for some, but for some it's, for me, it's critical.
(15:25)
So we have to make sure that the same diligence that we employ in protecting the government network, that we can also share that information out to help protect the corporate networks and those that operate that critical infrastructure. So we created a group that was focused on bringing industry together so that we can talk about the threat landscape as we saw it. So from a government perspective, we have a lot of access to information around what's happening both on a national, international level, and also when it comes to collaborating between industries. It's possible that something can appear in the financial industry that might move into the food and agriculture industry, but because those sectors aren't necessarily connected from a cyber perspective, they may not know that that's happening. So by bringing a committee together of those that operate critical infrastructure in the city and in the state and saying, here's the current threat landscape as we see it, here are all the risks and the things that you should be patching against. It gives us the opportunity, one, to bring information to the people that need it so they can do something about it. But it also humanizes the process for reporting when something goes wrong. Because if we're all in a conversation around how we protect the state, when something goes wrong, if you know that you have support that comes behind it, you're less likely to withhold that information and cause more damage to the infrastructure that may be compromised.
Penny Crosman (16:47):
Can you build sort of a dashboard of the whole city and where you could see threats happening in all different kinds of businesses and agencies, and that way you could start to see patterns like, this thing is happening over here and it's heading this way or this pattern. You might recognize it somewhere else and start to see, well, that's a particular kind of attack.
Matt Fraser (17:09):
So what's very interesting about what you just described is pretty much a recant of how we do physical security. So across the city right now, we have lots of partnerships with big financial institutions, big real estate institutions, those that have a lot of assets, right? Assets like sensory assets, things like security cameras and things along that nature. And we have strong partnerships in those areas so that we can see as crime happens and the specific crime that's plaguing certain areas, if we see a pattern of behavior, we can inform our policing strategy across the city to make sure that we can deter that where possible. Now on the cyber perspective, we're looking and building a construct so we can do that across the entire state. So we can look at threats as they come in upstate downstate, and then we can say, here are the things that we've done to disrupt those threats. And here's what you can do. For those of you that are part of the collective, and those of you who are not, here are the steps that you can take to help protect yourself. But what you just described is what we're looking to do now.
Penny Crosman (18:15):
It's interesting because it's often specific gangs that are doing multiple kinds of attacks. If you could find them and follow them around and watch what they're doing, I'm sure you could catch some more easily. What about things like access to city benefits like childcare, food stamps, affordable housing, et cetera? Are you using technology to make that access more available?
Matt Fraser (18:42):
For sure. So at the beginning of the year, one of the things that we took a look at, and beginning of 202 I should say, we took a look at where the city stood from a benefits perspective and the mayor ran on a platform that he was very adamant about democratizing access to city services. So some of the challenges that we had was that a lot of city services were fractured or fragmented between agencies. One such service is something like childcare. When you wanted childcare benefits at the start of the administration, you had a paper form that was issued by the state that had to go to one of three different agencies to see if you're eligible to one of those three childcare subsidiary programs. The problem with the process was that it was all paper-based. You had an SLA that basically said within 30 days you'll hear something via post mail, whether you were eligible or not.
(19:37)
And then the third thing was that if you only applied to one and not the other two, it would take you up to 90 days because you had to apply to each one individually to figure out where you were eligible. So to get around that inefficiency, we focused on the mayor's commitment around MyCity, one application that will essentially consolidate city services for our constituents, see where do we start? So at that time, in 2022, we're still in the middle of covid. We're looking for people to return to the city, to return to work. But one of the things that's a challenge, especially for those that need access to resources but don't have it, is childcare, right? No one is going to go to work unless they know that their child is safe. And getting access to those benefits would alleviate some pressure so parents feel comfortable going back.
(20:27)
So when we looked at what we did at MyCity, it became a very clear direction for us. Our North Star was like, let's get childcare sorted. So the first instance of MyCity was focused on building a childcare application portal where you could apply to those three programs in one place. You could get a pre-submission eligibility determination to say basic on the information that you provided. You're eligible for the program and you're eligible for this form of the program. You can make a more informed decision and apply for the thing that you were actually looking for. So technology now has removed a lot of the barriers that the paper-based process had and has created a lot of efficiencies. Now that's nice in the childcare space, but now how do we take that and expand it a bit further to cover more things? So we're now looking at how we can treat city benefits, like I mentioned this earlier, like credit card approvals.
(21:25)
At home, you'll get a notice in the mail that you're pre-approved for this card. Here's your pre-approved limit. Being able to do something very similar for public benefit, to send those that are recipient of one benefit benefit to say, hey, you are also pre-approved for these other benefits. You just have to complete the application process. It gives us the capability to put more resources in the hands of those that need it and remove the stigma with applying or appearing before government. And in that, it brings up the entire quality of life of the city, because if we can focus on those that have access to the lease and we can bring up the quality of life for them, the rest of the city will also move in that same direction.
Penny Crosman (22:10):
And are you using AI to help make some of those decisions about who's eligible and who's not, or do you think you might in the future?
Matt Fraser (22:17):
So that's exactly where we're looking at employing AI. We have an office that's focused on algorithms management and policy to make sure that anywhere where we put automated decision-making that we have the capability of assessing bias. Is there any bias that's implied by leveraging the process, setting up the guardrails, what decisions will the technology make? Which ones will it not be able to make? And making sure that thresholds are very clear where human intervention is required. So we are looking on a much broader scale to employ AI, to help us make smarter decisions, faster decisions, and remove the amount of time it takes for someone to get benefit from requesting that benefit.
Penny Crosman (22:59):
So we talked about how people are now back to work and we're seeing it. Companies all over the city are mandating anywhere from three to five days a week back to work. Does that affect your job -- more people commuting, more people in the city every day? Or are you already kind of equipped to handle that change?
Matt Fraser (23:21):
No, so we're already equipped for that change and we're ready for more and more people to come back to the city, right? It's like don't alleviate the pressure that more people come in. And I think for us, making sure that services are continuously available and reliable for those that commute in or live in the city is part of our core goal. Whether we have our native 8 million population or we have a surge of people that are coming in for an event, New York City can't be in a position where it stops. So we make sure that we do our best to ensure that everything keeps going.
Penny Crosman (23:55):
Do you have any sort of technology priorities for New York City for the next three to five years?
Matt Fraser (24:02):
I think for us, our tech priorities are focused on making sure that we bring services to our constituency at a very basic level. The mayor is appointed by the people, I am appointed by the mayor, and everyone that works in these organizations serves at a very basic level, the interests of the people. So for technology, we're focused on making sure that technology serves the entities that it's supposed to, which are the people of the city. So more stuff around benefits, access, more stuff around customer service when someone calls the government for support, being able to reduce the amount of wait time and friction that's in some of these processes and some other areas from a business and economic perspective, continue to make New York City the tech hub for the world. For the first time ever this year, New York City unseated Silicon Valley in terms of the number of tech companies that were started and growing in the city. So it's a good milestone to have, and it's no surprise that it happened under this mayoral administration because the mayor clearly backs tech. And for us, we want to ensure that that kind of growth continues to grow here and the community, the tech community in New York City is vibrant.
Penny Crosman (25:17):
And is there anything your organization can offer as sort of a carrot or incentive?
Matt Fraser (25:22):
We're working on a number of programs that are focused on both digital literacy and upskilling. So New York City, for those that may not be familiar, is the 12th largest economy globally. So if we were a nation, we'd be the 12th largest financial power. Somehow we managed to maintain that position where when over 40% of those that lived in our public housing community didn't have access to broadband, by bringing up the entire quality of life of the city and bringing up the skills of the workforce, we would have the capability to not only be the cultural source of the world, but we'd also have the best tech talent in the world here by building more of these programs here locally. So my office is focused on that and also democratizing the process for tech companies to get started in the city.
Penny Crosman (26:12):
That makes sense. What about the people who struggle the most? People who live under bridges and sleep in shelters, they don't have smartphones, they don't have laptops. Is there any way AI or any other kind of technology can be used to help them, raise them up?
Matt Fraser (26:31):
So we focus very diligently on bringing programs to, I say to market, but bringing programs to the city that has the capability of bringing those that have had access to the lease up. So I spoke about some of what we did around bringing broadband into the public housing community, but that's a "so what" sort of moment. It's like now they have connectivity. That's what now, if you look at one of those old proverbs, you teach a person how to fish, they'll eat for a lifetime, you give 'em a fish, they'll eat for a day. Providing someone with broadband is like the equivalent of giving them access to water. A device is like a fishing pole, and the knowledge you need is like the skills of how to use that pole to feed yourself. So we've done a lot of work around bringing people to water.
(27:21)
We have to do more work around providing access to devices and teaching them how to fish. And that's what in the next year or so, we've got a couple of programs that will come out that will talk about how you go from apprenticeship building skills and actually going into the city's workforce and actually getting a job. And I think that's some of the areas that we want to focus on to transition. Parts of the community that you've just mentioned are disenfranchised, that may be living on the street, that may be in shelters, but they don't have a pathway to financial freedom. We want to help them get there. But in addition to helping the folks outside, we also have to make sure that as technology evolves, our internal workforce can grow at the same rate that technology is growing, which is one of the reasons cybersecurity is one of the biggest skills that are in demand. Earlier this year, we launched our cyber academy. Now our cyber academy is focused on building cyber skills across the city's workforce to ensure that in every agency there's at least one person that's skilled at an advanced level on how to deal with cybersecurity threats. Taking programs like that and making that available to the general public will not only build skills, but it will break that poverty cycle because it'll give them pathways to employment that will be more than three to four times the average household income where they come from.
Penny Crosman (28:44):
That sounds good. Well, I'm afraid we're out of time, but Matthew Fraser, thank you so much for joining us today. It's been really interesting.
Matt Fraser (28:50):
Thank you very much for having me. Pleasure.
