Federal chief information officer Alex Benay announced he was leaving his position last week to become chief client officer at MindBridge Analytics, an Ottawa-based company that makes artificial intelligence (AI)-based auditing tools. His departure comes two and a half years into a role that began with fanfare and a mandate, from then-Treasury Board president Scott Brison, to “disrupt” the way government uses technology.
Benay is leaving as some of the biggest projects he’s spearheaded are about to be implemented. These include the announcement of which of three finalists will build a public-service payroll system to replace the over-budget and error-causing Phoenix. And, by September, departments will need to have a data strategy in place, under a plan that Benay was instrumental in shaping.
Ottawa’s top tech executive has maintained a larger-than-usual public presence for a senior public servant since being appointed in April 2017, frequently tweeting and speaking at conferences about the need for a digital overhaul of the government.
In a wide-ranging interview on his second-last day on the job, Benay graded the success of some of his major initiatives. He also discussed why he’s leaving the public service for an AI company, why turnover in his old job is a good thing and why he’s still hopeful about digital transformation of government in Canada.
In an exit interview with The Logic, Alex Benay, Canada’s outgoing chief information officer, discusses the progress of some of his biggest projects, like a faster, less expensive way for the government to buy technology; a plan for how departments should use the data they collect; and efforts to make Ottawa more transparent by releasing information to the public. The top federal tech executive is taking a job at MindBridge Analytics, an Ottawa-based firm that makes artificial intelligence-based auditing tools.
The conversation has been edited for length and clarity.
The most high-profile of your projects has been a new way that the government acquires technology to replace the approach of big contracts with long timelines across every department and agency. How has that process of transformation been going?
What we hopefully will have proven with the Phoenix replacement is that you don’t have to change procurement policy, because we didn’t have to do any of that. You don’t have to change the laws. You just have to change how you approach a problem.
Procurement historically has been “Here’s the requirements. Tell us your best solution.” But you’re defining the solution, as opposed to letting really smart people help you understand that maybe there’s a different solution altogether that you never thought of.
With Next Gen [the new payroll system], we went to the street with the problem. We then produced a series of [stages] that were user-tested the whole way through and that influenced the elimination process. And, we were able to do it in less than a year. Now we have three solid technology options to choose from.
How much adoption have you seen of that kind of mentality and this new agile process in other departments that are making procurement decisions about tech?
We’ve seen, for example, Shared Services Canada use an agile procurement process for the Canadian Digital Exchange Platform. We’ve seen Transport Canada use it. The biggest challenge in government—not just the Government of Canada, but other governments—is changing the process and the habits of people. It’s very rarely a policy or a legal issue.
So that just takes time and persistence and education. And so, those things are all underway now. In order to, as a whole, shift the process, we’d have to put a lot of people on the problem. We just don’t have that many people. So we’re kind of chipping away. I would give it a solid B.
Opening up data and information to the public has been an interest of yours for a long time. For example, you made 80 per cent of the content at the Canada Science and Technology Museum [Ed: where he was CEO] public. How are the efforts to make more information public going?
We’ve moved up in the ranks in the Open Data Barometer—we’re tied for first with the U.K. Now, my old hockey-competition blood is very interested in making sure we’re alone at the top of the ranking next time around. But it’s because we’ve been able to release more content. Treasury Board has a group now that is opening up by default the open-government team itself. And, other departments are starting to look to follow suit.
We were having a very active conversation around open science. Personally, I’m of the belief that heritage and science that are publicly funded should be done outside the firewall. The pace of adoption is good. I think where we have a gap is other countries seem to be able to commercialize the open data movement more than we can.
Flight trackers or Ancestry.com are billion-dollar companies, but for some reason, we’re not creating billion-dollar companies on the back of open government. The reason for that can be a million things. Maybe it’s corporate culture that’s not ready to interact with government; maybe it’s government not ready to interact with corporations. Maybe we don’t feed our startup ecosystem.
Another of your big projects was the roadmap for how the government should use its data. Departments are supposed to submit their data strategies in September, and that’s the first big milestone. Are you satisfied with the uptake?
What we’re seeing in departments is the realization that data is a big deal. The data strategy has helped cement that. What we’re not seeing is: how do we execute against it? For example, most of our laws were written pre-internet. That’s the next frontier: are our policies equipped for a digital economy? Are our organizational structures equipped for a digital and data-driven economy? I suspect the answer, often, is no. But to be fair to Canada, every public service and every government in the world is having a hard time with it.
For example, in order to get to a different service environment with citizens, we would have to change well over 150 pieces of legislation to enable data sharing across departments right now. Data sharing across departments is often seen as a risk to privacy, but actually, it’s not sharing data that’s the risk. [It] means institutions collect the same piece of information on you sometimes multiple times within a single department, and it increases the risk of breaches, of cyber hacks.
Internal documents that The Logic obtained via access-to-information requests show that some departments are concerned about the resources required for the data plans you laid out.
I actually fundamentally disagree when people say, “To do the data thing is more expensive.” To do the data thing is actually cheaper. We just have to change the business model around it. So when we did open by default at the museum, I think it cost us $25,000, reusing the same kind of tools that [another organization had] already bought. And that was a tiny little Crown corporation that had no budget. That institution went digital—there’s a documentary on it, actually, that TVO published. We just changed how people work.
We’ve done a Gartner comparative study against us versus other countries of how much we spend in digital. And, there’s a lot of areas where we overspend because we take an industrial-age design approach to a digital problem. When you release the data, and then you engage with stakeholders to create an outcome that doesn’t have to be controlled by government, you get faster, cheaper and better results. Through releasing data, through engaging publicly, we were able to come up with three viable technological replacements for Phoenix in one year, not two, and with $8 million and not $16 [million].
You’ve talked about moving more government applications to the cloud, and using commercial services to store government data. What’s needed to accelerate that?
If you look at it from software-as-a-service [SaaS], you’re seeing a lot of movement. You’re seeing a border services project [an online freight manifest] moved to software-as-a service. The Phoenix replacement is software-as-a-service. People are now [using] readily available products that can be customized a million different ways and that have been deployed around the world to meet our needs. That’s great, because it reduces the pressure on our infrastructure, which is aging. As far as [moving from data centres to the cloud], we have 71 active projects on the unclassified cloud service that Shared Services Canada offers over the last 18 months. So that’s a good sign.
Other countries do a better job of helping their startups and scale-ups into government spend.
Keep in mind, 300,000 people [work in the federal public service] and 25,000 people that work in tech and data in government—it is the largest operation in the country. So, it’s not going to shift overnight. I feel really good about the progress in the last, I’d say, 12 months. It’s only going to accelerate. The money is there to do it. And then the contracts are almost all there to do it, as well.
That’s one of the things I’m proudest of. When we got here, cloud wasn’t even an option. We had all these myths around security and privacy when, frankly, Microsoft and Amazon and all of them invest more in cybersecurity than most countries. We had to kind of bust all of those myths for a year.
I’ve heard from industry sources a few times that the issuing of contracts has been slower than they anticipated. And, the government’s public procurement database only lists 97 contracts that involved cloud specifically.
Anybody who says it’s not moving doesn’t have the right information. Of course, I would like it to move faster. But these are the largest operations in the country; you can’t do it in a month or two. And, anybody from the private sector saying we could, has never had to run something this big, that would be my opinion. We’re also dealing with the lives of Canadians. We fail, the border shuts down. We fail, people don’t get on a flight. We fail, your taxes are all screwed up. There’s a different level of risk. The pace is commensurate to the risk. And there’s been quite a bit of progress.
In August 2017, you wrote a memo about government tech buying about a study your office ran showing “a large concentration … across a small number of large international companies.” In June 2018, my colleague Zane reported that, for example, IBM got $1.9 billion in contracts after Phoenix launched in February 2016. Does federal tech procurement still favor big foreign firms? Or has that shifted in favor of innovative Canadian companies in your time?
Listen, IBM’s in every country. Microsoft is in every country. You could never get rid of them, and I don’t think anybody would want to get rid of them. That being said, other countries do a better job of helping their startups and scale-ups into government spend—the U.K., for example. We’re starting to shift to a more open-based procurement approach.
If you look at the AI vendor list [of companies allowed to bid on government projects], for example, 50 per cent of the 73 vendors on that list are Canadian [Ed: the current number is 62 per cent of 78 firms]. That’s a big deal. That is Public Service and Procurement Canada writing a procurement that is open-ended and that is directly related to investments made by [Innovation, Science and Economic Development] in this space a few years earlier that resulted in the creation of startups that are now part of a government supply agreement. And they’re not just all in Ottawa, which has historically also been a problem. They’re in Vancouver, Edmonton, Toronto, Ottawa [and] Montreal. That, to me, is a perfect example of, you know what, we can do this thing a little bit differently.
And in one of the most forward technologies out there.
Absolutely. Canada missed the boat, as a country, with cloud, seven or eight years ago. We had some Canadian suppliers with cloud. It’s unfortunate, because the data play and the cloud play for Canada could have been amazing, globally. We could have been the Switzerland of data. But we did not do that. And with AI, I don’t think we’re missing that boat right now.
Speaking of AI—why MindBridge?
Canada, specifically on the ethical AI front, has created, to my knowledge—and I’ve checked with a lot of global AI conferences and places—the only ethical guardrails in an organization in any sector in the world. So the Government of Canada has done that well, in collaboration with industry. The reason [for] MindBridge is the mandate to do good. The mandate to catch fraud, and to do it in seconds—not months later, after the fact—could change the entire financial ecosystem of the world.
Another big reason for me, personally, was that it was in Ottawa. I believe in the Ottawa tech scene. And my family’s here. So for me, moving to New York, or Qatar, even—I got some interest—just didn’t feel the same as growing out of Canada and Ottawa. And personally, I just miss growing something.
I had given myself two to three years in this current role. I don’t think you can do more than three years in this role if you’re actively trying to change things. There’s too many stakeholders; there’s too many things [that] could go wrong. There’s aging technology. It’s an exhausting role. But it’s probably the most impactful role you can have. So I needed to find something that had [an] equal impact, and a vision that has an equal impact.
Your predecessor, John Messina, also left after about two years. But it sounds like you don’t take that as high turnover.
If you [do more than] three years of actively changing the system, you’ll start making mistakes, you’ll take things too personally. It’s hard. Not to be too personal, but my wife has worked in tech for a long time. For a period of time, it was hard for her to switch jobs, because [of] perceived conflict. I don’t do procurement, but nevertheless, it has an impact on your family.
It’s just a different lifestyle, you have to be ready for that. It also comes with the territory. Like we chose to make full transparency of what we were doing a priority, which means that comes with increased scrutiny. We could have hid behind the desk. But you don’t do change from behind the desk, you gotta do change out in public. And that makes it harder. I don’t think that two to three years of turnover in a tech role [is high]—I think three or four [years] is slow.
If you’re leaving a list to your successor the new minister of digital government after the October election, what’s on it? Is there an amount of money? Is there a long list of legislative changes?
In a digital world, your laws will have to change all the time, regardless of country, regardless of political system—think Uber, Airbnb. Your regs and your laws have to change quickly. It’s not about the money, it’s about the approach. We don’t need a large investment on digital identity in this country, for example, we need a federated digital identity model [connecting the federal and provincial governments], because we live in a confederation. We need to look at how we manage the digital economy infrastructure, holistically as a nation. We need to look at machine-to-machine communications for transactions that are safer, more private, more protected and faster and create more wealth for the country.
That’s the kind of area that if I was still in the role for a couple of years, we’d be weighing [in] very heavily on. That’s the kind of stuff that touches citizens [and] business; it helps for competitiveness as a nation. Other countries are moving aggressively in this area, whether it’s Denmark, or Estonia that’s been there for a long time or South Korea.
There is this thing called the digital revolution and digital economy, and we’re responsible for shepherding people through it.
And that is a role for the CIO of the Government of Canada?
I think it’s part of the role of the CIO of Canada. I think it’s part of the role of half a dozen departments in the federal government. I think it’s part of the role of provincial governments. I think municipal governments are probably the ones that are feeling this the most right now. I’m not going to get into this, but if you look at Sidewalk Labs, and you look at the Internet of Things and connected cities, there’s a lot of things that we have to give a hard look at. Privacy in a digital age is not privacy 15 years ago. What’s the intersect between cyber service and privacy and convenience? All of these things are fundamental societal questions.
Share the full article!Send to a friend
Thanks for sharing!
You have shared 5 articles this month and reached the maximum amount of shares available.Close
This account has reached its share limit.
If you would like to purchase a sharing license please contact The Logic support at [email protected].Close
Share the full article!
Share the full article with your friends. Recipients will be able to read the full text of the article after submitting their email address. They will not have access to other articles or subscriber benefits.
You have shared 0 article(s) this month and have 5 remaining.
Do you think that Canadian governments are suffering from a bit of a tech hangover— either in perception or in sort of risk tolerance—after projects like Ontario’s failed eHealth medical records system, or Phoenix? Has that made people less ambitious, or made the public more skeptical of these digital transformations?
I think that’s probably fair. A lot of large failures have happened in this space. Large failures happen in the private sector; they’re just not reported. I’ve worked in both, and I can tell you that that’s the case. With Next Gen, for example, the whole project was public. There was political support to make it public. The documents are mostly available online. Every small step was communicated. That means you’re building confidence as you’re going through the process.
The other part of that is it was step by step. It wasn’t trying to boil the ocean and do a big launch. The days of the big mega-project have to end. You [have] to be in a continuous state of tinkering. Two small steps forward, maybe a step back—[but] it’s small; it’s not catastrophic. With [the new payroll system], they’ve been able to prove they can do big things in less than a year and be very communicative, be very transparent. Tens of thousands of public servants were engaged in the process; it was was very user-driven.
In May 2017, you said, “If we don’t change how we do things, we’ll potentially become more and more obsolete. It’s a big gap and it’s growing.” Do you still believe digital transformation of government is possible in Canada?
Oh, absolutely. What Canada is not good at is telling its story. We were the first country, decades ago to [create a] a tax service ecosystem, where Intuit and TurboTax and other [tax filing software] companies were able to flourish. [At] Veterans Affairs Canada, [the] pensions for life [online platform] was done in 14 months, from budget announcement to delivery—fully agile, open source, consultative with veterans. And, it was a great success. Next Gen—amazing work. The AI work that the ethics team did—amazing, like world-leading stuff.
It’s totally possible and, in fact, we’ve been doing it. We just got to get better at telling the story.