0:00 George McEwan ｜ The Landscape of SEDI ｜ KERI Conference 2026
0:02 So everyone, I'm George McEwan. I'm the
0:04 privacy architect for the state of Utah.
0:06 So there's a lot of synchronicity that
0:08 happens when you go to these various
0:10 events. And amazingly enough, I'm going
0:12 to talk a little bit about healthcare in
0:14 the rural healthcare setting and why
0:15 we look at it the way we do. One of
0:18 the things I'm going to make very clear
0:19 is that the state is going to declare
0:20 some boundaries in what it does in the
0:23 KERI infrastructure and what the
0:24 expectations are for the community
0:27 because one of the things we learned as
0:29 I was the IT director during the COVID
0:32 era at the department of health, there
0:34 was definitely some resistance to where
0:36 government had offered the technologies
0:39 or applications or worked with platform
0:42 providers to provide things like
0:44 exposure notification and the backlash
0:47 was very intense. That's
0:50 probably the best word I can use. And so
0:52 we realize that where government can
0:54 help and we can set regulations and we
0:56 can set frameworks in play, there's
0:58 definitely a spot where the government
0:59 has an edge and we say this is private
1:01 sector and for the individual to handle.
1:04 So, what I'm going to do is I'm going to
1:05 run you through a little bit of a
1:06 presentation that we do with we'll
1:09 call them non-nerds, normal people.
1:12 Because we can talk in shorthand in the
1:13 SEDI community and in the KERI
1:15 community and cyber security. We have
1:17 words they don't understand. So, what
1:19 we did is we put together kind of a
1:20 generic presentation that we add slides
1:23 to depending on who we're talking to.
1:25 So, realistically, we hit them on these
1:28 points. Digital identity is here. It's
1:30 coming. It's going to be something. Now,
1:33 one of the, I would say, the beauties of
1:34 the SEDI program is Chris Bramwell
1:37 said, "Well, we could take this and
1:39 put this against what are the
1:41 constitutional rights that someone has."
1:43 And by doing so, you get past all of the
1:46 vendor needs to make money. You get you
1:48 get right down to the root of the
1:50 problem, which is there needs to be
1:51 accountability and that your rights need
1:54 to be preserved and that privacy is not
1:56 impossible when you are talking about
1:59 digitally issued credentials. And of
2:01 course then we get into the stakes.
2:04 If the government doesn't choose or
2:07 regulate, the market does it itself. And
2:09 the goals of a open market are a
2:13 little bit different than what a
2:14 government needs to be providing the
2:16 services that a citizen would want. And
2:18 of course, we know that besides
2:20 governments, there's also organizations
2:22 that put together specifications that
2:24 may or may not be rooted in the same
2:26 value system that we're trying to
2:27 maintain. So we're very adamant that we
2:30 want to make sure that whatever we do in
2:33 the SEDI environment, it's reflective of
2:35 the values, goals, and rights that are
2:37 held here in America. Now, it is not
2:40 hard for this statement to be believed
2:43 by all of you, right? Whoever has the
2:46 data has the control. And the model
2:49 right now, everything to the left or
2:51 sorry, my right everything over there
2:54 is how we work today. Your identity is
2:57 owned by other individuals. It's granted
3:00 back to you with limited rights. And
3:02 then if you've ever looked through all
3:04 those long user licenses, in effect,
3:08 you've given up all control for the
3:10 services that you have. And
3:12 this isn't limited to just, hey, I got
3:14 myself an email address for free. Your
3:17 financial institutions do this, your
3:19 hospitals and your HMOs, your care
3:22 organizations, they're holding a lot of
3:24 the data that belongs to you. And
3:26 because it's set up that way, you are
3:29 basically powerless to do anything about
3:31 it. And in a lot of cases, you'll find
3:33 that incorrect data may be used in
3:36 profiling you to create an improper
3:39 picture of who you are.
3:42 So really what we want to accomplish
3:44 inside the SEDI framework is the idea
3:46 that you are the individual who controls
3:49 that information because
3:52 wherever the information sits, it is fair
3:54 to say without contradiction that
3:57 impacts the financial outcomes. It
4:00 impacts your opportunities. There are a
4:02 lot of things that get impacted because
4:04 you don't control what should
4:06 rightfully be yours. And this is not to
4:09 say that everyone did it for a nefarious
4:11 purpose. I mean, profit-seeking motive.
4:13 We're a capitalist society. Yeah, that's
4:15 understood. But the problem is we built
4:17 everything to be so exploitative of you
4:20 that we forgot that you are the
4:22 individual who ultimately gets hurt if
4:24 something goes wrong. In the state of
4:26 Utah, inside government,
4:28 there's a very good way that we look at
4:29 data. When we make a data error in
4:32 government, it's a negative outcome for
4:34 a real person. Classic example, oh the
4:38 Medicaid eligibility system, which is a
4:40 real time system, goes down for a couple
4:42 hours. Now, if you're standard IT,
4:44 you're like, big deal. Things happen.
4:45 We'll get it back up. No big deal. But
4:47 then when the Medicaid director marches
4:49 in with their flesh hook and rips a
4:52 pound off of you and says, "Right now,
4:54 somewhere in Utah, there is a single
4:56 mother who took two buses two hours out
4:59 of their day to go to the pharmacy to
5:02 get medication for pain management." and
5:04 the pharmacy directors at that
5:07 facility said, "We can give you one pill
5:09 because we can't confirm your
5:10 eligibility. You'll need to come back."
5:13 That's the sort of situation we deal
5:14 with when we talk about our data errors.
5:16 And of course, we try very hard in
5:18 government to be good stewards of that
5:20 data. But realistically, if the
5:23 individual has a lot more control, who's
5:25 a better advocate than you? You are
5:27 definitely going to be your best
5:28 advocate.
5:30 Now, when we get into the concepts of
5:32 state endorsed digital identity, we
5:34 really try to keep this as simple as
5:36 possible. And this is probably
5:39 way too high. And everyone's like,
5:40 "Yeah, I get this slide immediately."
5:41 But what we try and explain to them is
5:43 those core concepts that we put into the
5:45 bill of rights that identity is yours.
5:49 It's not issued by the state, but the
5:51 state has a role to endorse it for
5:52 certain legal purposes. And of course,
5:54 we get into the concept of what is a
5:57 digital identity? How does it get
5:59 created a little bit and then we go into
6:01 the whole idea of how it becomes your
6:02 state endorsed digital identity through
6:04 this process of vetting. Now there are
6:07 some potentially debatable questions on
6:10 here. We don't get into the politics
6:12 of this, but we do know that there are
6:14 some markers that need to uniquely
6:17 identify you to be part of that digital
6:19 identity. And this isn't to say that
6:21 these are forever attributes. For
6:23 example, your address, your biometric of
6:27 your face, those are going to change.
6:28 And we have existing laws that handle
6:30 that. For example, if you are
6:32 getting a driver's license at 16, that's
6:35 not valid for more than 5 years because
6:37 we know you're going to change by the
6:38 time you get to college. And of course,
6:40 with the residents here in Utah who have
6:43 driver's licenses who are over 18, it's
6:46 a 10-year window. So there are
6:47 opportunities for those things to be
6:49 changed as part of the credential or to ..,
6:52 for lack of a better term, lose the
6:53 endorsement.
6:55 Now when we talk about healthcare data
6:58 there are a ton of different places
6:59 where this applies. Now we had
7:02 healthKERI talking about the provider
7:03 payer relationship. That's just one of a
7:06 myriad of relationships that exist
7:09 between you and your healthcare
7:11 interactions. Whether that's the
7:12 ownership of your individual data, your
7:15 desire to participate in some sort of
7:17 demographic mapping. Today, we
7:19 don't do that. Today we have systems
7:21 that basically, at government control,
7:24 take your data and aggregate it for
7:26 research purposes.
7:29 And for some people, that's not
7:31 shocking. But when you go to a
7:33 state and you say, "Hey, do you have an
7:34 All-Payer Claims Database?" If the
7:37 person's in the no, that means that
7:39 every encounter you have that's part of
7:41 your insurance program in that state
7:44 with your physician and the notes
7:46 associated with all your ICD-10 codes,
7:48 everything that happened during that
7:50 medical encounter is now in a government
7:52 database for the purposes of looking at
7:56 healthcare outcomes. Did patient A and
7:59 patient B who received different
8:00 medications, which one had the better
8:02 outcome? So there's a lot of good
8:04 intentions in terms of why these data
8:06 sets exist. But I will also tell you the
8:09 most dangerous thing in government is a
8:11 well-meaning bureaucrat. It's not that
8:13 they're trying to do something
8:14 nefarious, but sometimes they try and
8:15 get hold of data and put it together to
8:18 create a picture. Well, where you don't
8:20 have any interface to this and you don't
8:22 have any control over that data, there's
8:24 the possibility that that inference,
8:26 that conclusion may help you, it may
8:28 harm you. So this is part of the reason
8:30 we want to start looking at how we
8:32 switch this model around so that you
8:34 have more executive control and how your
8:37 digital twin story gets told.
8:40 Now when we talk to them about the state
8:43 endorsed digital identity, we also like
8:45 to help people understand this
8:46 concept of perpetual identity. And this
8:49 is where we differ with the mobile
8:51 driver's license community and they're
8:53 saying you can use it for everything.
8:54 I'm like, you can use it for everything
8:56 that you don't mind having burned down
8:58 if you move to a new state. That's
8:59 totally fine. But in our particular
9:02 instance, we want to be able to say that
9:04 there are certain documents that you may
9:06 associate with your SEDI and they
9:09 could be your medical records, you know,
9:11 your digital driver's license. Any of
9:13 these items that you have associated
9:15 with your SEDI can be created and you
9:18 may have information that you are going
9:20 to selfassociate like, hey, here's my
9:22 claims history. Here's some of my MRI
9:25 scans, things that you have in your
9:27 possession that you want to get into
9:28 your digital records but currently
9:31 don't have. And this is where these
9:32 functionalities like Blue Button are
9:34 fantastic is you have the ability to say
9:37 this is categorically my MRI. I'm going
9:39 to send it to you electronically. You
9:41 can pull it directly into my
9:43 record set. That's the level of control
9:45 that we want to have. Not: go to doctor
9:48 office A, pay $50 for a doc fee to have
9:51 them really, it still cracks me up, fax
9:54 information to doctor B, and then
9:57 hopefully they didn't lose anything in
9:58 translation and the whole packet's
10:00 complete, which you generally find out
10:02 on your next medical appointment that
10:04 it's not somebody dropped pages 8
10:06 through 12, and now you got to go
10:07 through this whole cycle again. So, we
10:09 have these inefficiencies that are in
10:10 the system the way we run it today. And
10:12 we're telling people, no, if once you're
10:14 in control of this, you can start to
10:16 aggregate your data as necessary. And of
10:19 course, we talk about the concept of
10:20 perpetual ownership. If I move from Utah
10:23 to Texas, there are certain privileges
10:25 that go away. I don't keep a Utah
10:28 driver's endorsement anymore. I have to
10:30 go get a Texas one. But everything else
10:32 that I had associated with my SEDI is
10:35 still valid. And in fact, because I have
10:38 a SEDI that's authorized by Utah or
10:40 endorsed by Utah, Texas can say, "We
10:44 know that's Utah. So, we're going to
10:45 accept these baseline facts that you've
10:47 asserted. We can bring you right into
10:49 the system and we can move on." And hey,
10:51 guess what? You can go get your Texas
10:54 driver's license endorsement without all
10:56 the reproofs that we have to do along
10:58 the way.
11:00 So when we looked at SEDI as a
11:03 functional item, we wanted to build into
11:05 the framework some very important
11:07 principles. And of course you can see
11:09 them up here. We break them into some
11:10 general categories. Now this is where I
11:12 want you as KERI experts to start
11:15 ticking off in your mind which one of
11:17 these items are solved by KERI, which
11:20 one of these ones are actually enhanced
11:22 by KERI, and also when we get to the
11:24 end, I'm going to give you a little
11:26 here's what we're not going to do at the
11:27 state level. Okay? So privacy and
11:29 autonomy this is pretty straightforward.
11:32 We want you to have interactions with
11:34 your digital identity that the
11:36 government does not monitor has no
11:38 capability to and also third parties.
11:41 We do not want crosscontext correlation.
11:44 We want you to have individual
11:45 relationships and as such there are
11:48 technologies that will do that. Funny I
11:50 think there's one in the room here
11:51 that might do that. We do want you to
11:53 have this capability of selective
11:54 disclosure. I was telling Mike a
11:57 story where I went to my grandkid,
12:00 picked them up for a fun day out and we
12:03 went to a trampoline place and the first
12:05 thing they did was get grandpa to sign a
12:07 liability waiver. And as I was looking
12:10 at it, the first thing I thought is, you
12:12 do not need my birthday. You just need
12:15 to know that I'm old enough to enter
12:16 into a release of liability. And I kind
12:18 of sparked because not only would that
12:21 be a problem, I don't actually have
12:23 permission for my daughter to release
12:25 liability for her children. So there's
12:26 some other things that could get cleaned
12:28 up in that situation, but this is part
12:30 of the community education. This is the
12:32 open sourcing of privacy we hope to have
12:35 happen is I'm not going to have every
12:37 single interaction where I'm going to be
12:38 able to ask why, but we do want to get
12:40 people and just across the world saying,
12:43 "Do you really need that for this
12:45 transaction?" And that's part of where I
12:47 say in keeping with what Chris has
12:50 said, a 10-year journey. Now, if you
12:52 think back in cyber security, how long
12:55 did it get people to understand the
12:57 Nigerian prince scam before we could go
12:59 on to that's a fishing email and then
13:02 that became part of the lexicon that
13:04 people use? It took a long time. I
13:07 call it about a 15-year window just
13:09 going back over my history and saying,
13:11 "Yeah, it's about when we got people to
13:12 understand it." And still we fall for
13:14 these things. But this is the same
13:16 journey we're going to be on with
13:18 selective disclosure. We have to get
13:20 people understanding "No, no, no, you're
13:21 in the driver's seat. You don't have to
13:24 give up this information to be able to
13:26 receive services." Now, there's a major
13:28 medical provider that I'm not going to
13:30 mention them by name, but when you go in
13:32 to receive services, they give you this
13:34 great set of disclosures. And one of
13:36 them is basically a blanket that says
13:39 you're seating all your rights for all
13:41 your data that we may use in research
13:43 purposes including biological samples.
13:45 We don't have to notify you. And by the
13:47 way, if you don't agree to this, we're
13:49 not treating you. That is scary that
13:51 they have that capability. But as we get
13:54 a community that's becoming more savvy
13:56 about these things, not just small
13:57 groups like us, but the larger
13:59 community, we can affect change and make
14:01 those policies flat out illegal. Okay?
14:04 And of course, one of the things that I
14:06 do want to mention is personas in the
14:09 term that we're using it in is basically
14:11 helping people understand. You don't
14:13 have to stand there at the counter and
14:14 pick element A, element B, element.
14:17 There are certain transactions you're
14:19 going to have that are going to be
14:20 standard. If you go to
14:23 a bar or you might have a pub crawling a
14:25 persona that says, "Here's my name,
14:27 here's my face, and here's when I want
14:30 you to cut me off and call the Uber for
14:32 me." That's going to be something that
14:34 we would expect the wallets to start
14:36 supporting is how do I map these things
14:39 very quickly so I can get to through
14:41 the interaction as frictionally
14:43 frictionless as possible.
14:46 So we get the next one going. Okay, I
14:48 talked a little bit very fast about
14:50 delegation and guardianship the other
14:53 day. Probably faster than I should have,
14:55 but we'll talk about it some more.
14:59 Delegation and guardianship is one of
15:02 the most critical relationships that we
15:04 see both from a government and a
15:06 societal capability. From a government
15:09 standpoint, there are multiple programs
15:12 that operate based on the relationship
15:14 that a parent has to a child a
15:17 caregiver to a foster child. You can you
15:20 can go down the litany of them and
15:21 you'll find there's a lot. In the story
15:23 I told yesterday about getting a durable
15:26 medical power of attorney for my
15:28 father while he was basically passed out
15:30 on the bed. He doesn't really remember
15:32 that. And there are situations where he
15:35 may want to rescind that. How does he do
15:38 that? All we've got is a scroll on a
15:40 piece of paper that I can guarantee you
15:42 is not his signature, but it will hold
15:44 up in court because a notary came, took
15:46 400 bucks in hash from me and said,
15:47 "Yep, he had a positive
15:50 acknowledgement of what happened." So,
15:52 when we have these situations of where
15:54 we have vulnerable individuals who need
15:56 assistance or you have an aging adult
15:59 who quite frankly need you to help them
16:01 with their finances and other
16:03 items, it's really cumbersome today.
16:07 And I've had to do this for one relative
16:09 where we had to bring the court order
16:10 down to the bank and then they went
16:12 through this whole trust and
16:14 verification process. They called
16:16 the courts. They wanted the make
16:18 sure that the all the numbers matched
16:21 up on the case and everything. It was
16:23 ridiculous. And it and fortunately, they
16:26 only had one bank. I was thinking, what
16:28 if they had like an investment portfolio
16:30 completely separate and that investment
16:32 company isn't even in my state? I
16:35 got chills on how long that would take
16:37 me to help them out. And with the
16:40 technologies that KERI is providing,
16:42 there's the ability to really do that
16:44 quickly through these the these designed
16:46 and delegated guardian relationships.
16:48 And as I mentioned in that little
16:50 vignette yesterday that all my brothers
16:52 and sisters were sitting around the
16:54 table arguing,
16:56 I realized there's six adults here that
16:59 could come to some consensus on what is
17:02 best for the situation. And it would be
17:04 in the interest of my father or my
17:06 mother to have us have that sort of
17:08 capability. So multi- signature to be
17:10 able to authorize these activities.
17:13 Totally important. Now we talked a
17:15 little bit about agentic AI. It's
17:18 coming. I'm playing with a lot of
17:19 them. They're still in my little
17:21 sandbox, but some of them are getting
17:22 more useful. And one of the things that
17:25 I would find extremely useful is I was
17:28 playing around with Open Claw and I
17:30 decided I wanted to
17:31 implement Gemini's or Google CLI
17:34 and it's like well I usually ask Grock
17:37 how to do that but I thought I would ask
17:39 Gemini directly and it came up with this
17:41 big fat don't do that is a
17:43 violation of our policies and terms and
17:45 conditions and we will shut you down no
17:47 ifs ands or buts, you know, "Bad you!"
17:51 And I realized as I thought through what
17:54 would have been the implications of
17:55 losing my I call it my burner account,
17:59 but I have not dozens, I have hundreds
18:02 of logins that are tied to that.
18:04 Hundreds of logins. So if that company
18:07 decided to take it away from me, I would
18:09 break my digital life in more ways than
18:11 one. So
18:14 here's your product placement idea. If
18:15 anyone wants to add this, I would love
18:18 that once my SEDI goes live that I have
18:21 the capability to go and break all those
18:23 relationships automatically and
18:24 reassociate them with an identity that I
18:27 physically control. That's where I want
18:29 to end up because at the end of the day,
18:32 if we leave all those accounts and logins
18:35 out there, we're seeding our power to
18:37 control our own destiny. And of course,
18:39 we've talked about the permission and
18:41 consent the demo for KERI blocks.
18:44 some of you might have seen at IIW [#41] was
18:47 fantastic. For those who didn't, and
18:48 I'll just give you the two-minute
18:49 overview, the goals are the same.
18:52 Parents do not want their children in
18:54 environments where there are predators.
18:56 And we know that there are online
18:57 communities that despite their best
19:00 efforts, wink the profit
19:02 motive is still there. So, they're not
19:04 investing in the technologies that will
19:06 protect children. And in this particular
19:08 example, we had a parent credential,
19:11 a SEDI-based credential that did a
19:14 derivative credential for their child
19:17 and said, "This is my child. This they
19:19 are 13 and they can be on this system
19:22 from 5 to 7 PM, two nights a week,
19:25 and they can only talk to other children
19:27 who have a validated SEDI credential."
19:30 And it's fun to watch the chat window
19:31 and all the people disappear who are not
19:34 able to talk to your child. Now, some
19:36 people think that's overly
19:37 restrictive. You're going to have your
19:38 genius child get around it anyway. That
19:41 is where we do have that interface
19:43 between technology and law. Especially
19:46 in Utah, we've talked a lot about the
19:47 harms to children through social media
19:49 and other ways. I don't think it's
19:52 going to be too much of a stretch once
19:53 the technology is ubiquitous to have
19:56 the law catch up to help parents
19:59 manage their children's online
20:01 environments. Okay. Security and
20:03 recovery. One of the things that when I
20:06 read Sam's paper for the very first
20:08 time, I was like, "Okay, why is it that
20:11 we've been doing PKI this way for like
20:13 25 years at this point and nobody
20:16 figured out this really obvious thing,
20:18 which is if you can't tell what the key
20:19 state is and where you're going next,
20:21 you're back to that whole trust model
20:23 and all the things we did to make CA's
20:25 work and the intermediate certificate
20:27 and by the way, get rid of this one."
20:28 And I was like, Sam, it was genius how
20:31 simple it was when you looked at it, but
20:33 it was one of those things
20:35 that, until somebody came up with that
20:37 answer, we were going to still suffer
20:38 from the same problem. So where KERI
20:40 has done a lot of things to help us out
20:42 and of course like Sam and I have talked
20:44 about this over the last three years
20:45 about detection of compromise, the KERI
20:48 specification continues to get
20:52 updated to meet the real world
20:54 demands that we have. And one of those
20:56 is this whole idea of the mutual
20:58 authentication. We've just talked about
21:00 that with the phones. We've talked about
21:02 that with healthcare providers. And I
21:05 will predict something that is going to
21:07 be in the future. So we have health
21:08 information networks all across the
21:10 United States. And they make a lot of
21:12 money brokering the transaction or
21:14 uplifting it from one EDI format to
21:16 another. What a world it would be is if
21:18 I'm the provider and I can reach out
21:21 directly to a discovered payer
21:22 interface. we have this mutual
21:24 authentication and I can just move that
21:28 significantly changes the overhead that
21:31 a provider office has to have. I mean I
21:33 I know that my dentist for example he
21:36 has
21:38 about nine administrative personnel.
21:40 He's one dentist with two hygienists but
21:42 he does a lot of work in social
21:44 services programs and he has to have
21:46 that level of people to continue to keep
21:48 up with the paperwork. And of course
21:50 that's kind of reflected in his pricing.
21:52 So every couple years he goes off plan
21:54 and I have to pay more to use his
21:56 services. But if we could get those
21:58 administrative burdens lowered, think of
22:00 where we could go.
22:03 No credential works unless there's
22:05 comprehensive utility. One of the
22:07 long-standing jokes about the mobile
22:09 driver's license is that it works at the
22:12 TSA. Well, it doesn't. Their terminals
22:15 are down. They're like, "Well, hey, do
22:17 you have your physical license?" And
22:18 I've heard this one myself because that
22:20 interaction will go a lot faster than
22:22 trying to play around with the terminal.
22:23 So it doesn't have a lot of usability
22:26 and the models were wrong.
22:30 One of those models was is we tried to
22:32 monetize in the wrong way or at least
22:33 vendors did. In our case the
22:36 our we're an alcohol control state. So
22:39 you have to go to a state liquor store.
22:41 But when you are the director of the
22:43 state liquor store and a vendor comes to
22:44 you and says, "Hey, by the way, I need
22:45 you to pay X number of dollars per month
22:48 per store to be able to accept this
22:50 digital credential." And they're like,
22:51 "Wait a minute. Isn't that issued by
22:53 public safety?
22:55 Why do I pay money to public
22:58 safety or to a vendor to accept a
23:00 credential that the state issued?" And
23:02 of course, their answer was so logical.
23:04 Well, I can look at the I can look at
23:06 the state ID or I can look at the
23:08 driver's license for free. So adoption
23:11 is a problem. There are a numbers of
23:13 areas where we hope to see functionality
23:16 come along like I talked about replacing
23:18 username and passwords or taking over
23:20 accounts. We want to make sure that this
23:22 area which is where the ACDC comes in
23:24 great is offline verification. I can't
23:27 have my environment completely breaking
23:29 because I don't have online access at
23:31 all times. Now, Joe Jackson during our
23:34 our study panel, he talked about our
23:37 offhighway vehicle OV certificate
23:40 that we did. And when Joe and I talked
23:42 about it, I said, "Joe, this is fine for
23:44 X.509 technology. Completely fine. It's a
23:47 lifetime endorsement. It's never
23:49 revocable and you have to supply a
23:52 primary credential with it. So, just
23:54 because you show the QR code and it has
23:56 your information embedded in it, you
23:58 have to show a driver's license." So we
24:00 said, let's go down the road with this
24:02 one because we want to make sure that
24:05 there's an appetite for these
24:06 credentials and that it would be
24:08 extremely low risk. So Joe told you a
24:10 little bit about and things we've
24:12 learned along the way there. But I'll
24:15 tell you what, the funniest thing that
24:16 happened is once we release those, the
24:19 boating community, everyone who's owns a
24:21 boat, I've heard this story like 40
24:22 times now. They're like, I own a boat
24:25 and you guys have this OHV thing that I
24:27 could have on my phone. Why do I have to
24:29 have a printed PDF? And they always say
24:31 paper on my boat. We're in the water.
24:34 Why do I have to have paper? So, we kind
24:36 of smile and say things are coming. The
24:38 the appetite is there for these
24:40 credentials. We have no loss of ideas
24:43 and use cases of where we want them to
24:45 go. Key to the whole success is the open
24:49 standards, no vendor lockin.
24:52 We have been very clear to the community
24:55 that security by obscurity is never
24:58 going to be a solution we're ever going
24:59 to be interested. Somebody brings me a
25:01 closed source package and says it's
25:02 magic, I don't believe them. I want to
25:05 see under the hood and in some cases I
25:08 lack the time these days, but if I can
25:10 get down into the code and look at A
25:12 plus B does equal C, I get a lot more
25:15 comfortable. And I try to point out to
25:17 non-technical people, the reason Apache
25:20 is such a successful project and how has
25:23 the high level of security it does is
25:26 there's lots of people looking at it and
25:28 that's where we need to be in identity
25:30 systems and identity credentials because
25:33 I can tell you across this room and I
25:35 know some of you by face, some of you by
25:36 name, there is a lot of expertise that
25:39 can get into a room and say A and B are
25:42 not equaling C and here's why. And as a
25:44 state, I don't want to have to broker
25:47 broker that and be able to say, "Yeah,
25:49 we made that decision on our own. We
25:51 want the community to support us because
25:53 ultimately the beauty of this is, it's
25:56 your data." And by golly, you're going
25:58 to care about your data. So, I'm going
26:00 to get some really good feedback as we
26:02 go along in this program, which we have.
26:04 And we enjoy the fact that you can do
26:06 that. And as Chris has said, if a
26:08 standard requires you to pay to be able
26:10 to understand your interface between you
26:12 and your government or how your identity
26:14 works, that's a non-starter for us. We
26:16 just we just won't go there.
26:18 Now, this part about any app becoming a
26:20 wallet, what we mean by that is we do
26:23 have some digital credentials in the
26:24 state already. For example, if you
26:26 download the Department of Natural
26:28 Resources app, you can get a QR code for
26:30 your fishing license and your hunting
26:32 license. That's great. But that's only
26:34 inside their little tiny ecosystem. We
26:36 want to be able to broaden that out. But
26:37 that doesn't mean they have to throw
26:38 away their app because they've
26:40 invested some good money in it. They
26:41 also have some poaching features.
26:44 So you can take a picture of a carcass
26:45 and a GPS and it will send it to you DNR
26:48 enforcement officers. So we don't want
26:50 to say "No you got to go
26:51 through the wallet only," but there
26:54 are going to be instances where we will
26:56 see extension of SEDI credentials
26:58 inside existing applications in the
27:00 state. The other thing that we are
27:02 very aware of is the first mover
27:05 advantage for a lot of people is not
27:07 going to be going through the platforms.
27:09 They're going to be doing their own
27:10 trusted apps. To me, there are many
27:13 Roots of Trust that can come into your
27:15 relationships. For example, I may trust
27:17 my bank and you say, "Hey, we're going
27:20 to help you out with your Wells Fargo
27:22 wallet and we know about these
27:24 pre-key rotations and we provide a
27:26 service where you can escrow those keys
27:28 with us and using your multisig
27:30 capability. Us plus you plus whatever
27:33 you know key material you have, we can
27:36 unlock your next key for you." That's
27:38 fine and that's perfect. And I want to
27:40 harp on this just a little bit. The
27:42 state's credential is the SEDI. We are
27:45 not the wallet. We're not your key
27:47 escrow. We're not your key recovery. I
27:49 kind of smile and my wife says
27:51 this a lot. She's a children's
27:52 librarian. She works in a public
27:54 library. She has a little counter. It's
27:57 almost her personal drinking game where
27:59 she says, and I quote, "I'm sorry, I do
28:02 not know your Gmail password. I can't
28:04 help you with that. Do you have somebody
28:06 you can call to help you with that?" And
28:08 she has people who argue with her. No,
28:10 no, I used this computer yesterday. You
28:12 should know my password. And so from the
28:15 state, I suspect we're going to get
28:17 something like that where people are
28:18 going to be like, well, you issued it.
28:20 You should know how to recover my keys.
28:22 Nope. That's part of this whole
28:24 idea of, hey, I'm turning it over to
28:26 you. Now, with the wallets, one of the
28:28 things I do also want to say is that the
28:30 state definitely is going to be looking
28:32 for third parties who will go and audit
28:34 wallets and provide some we'll call
28:37 them seals of approval. What I
28:39 anticipate happening is when you go to
28:41 the state to get your
28:42 endorsement for your study is we're
28:45 expecting a cryptographic hash and a
28:46 signature of the application you're
28:48 using. And I can say "Yeah, that's
28:51 one of the ones we recognize as meeting
28:53 the standards of the state. We will go
28:54 ahead and finish up this issuance
28:57 process and give it to you."
28:59 Steve McCown's going to
29:00 take the source code directly and he's
29:02 going to compile his own wallet and make
29:04 sure it meets his requirements. So when
29:06 Steve comes to the state and says "Here's
29:09 my wallet!" and in the presentation he
29:11 provides us a hash and a signature we
29:13 don't know we'll say "Okay, Steve, we're
29:16 happy to issue to this to you but we
29:18 also want you to affirmatively consent
29:20 that the state doesn't know anything
29:22 about the security of what you're doing."
29:24 Yes, you want it. Here you go. Because
29:27 we do want that freedom of choice and
29:28 the ability for people to operate, but
29:30 we also want to have that
29:32 guardianship that the state has, which
29:34 is the public interest to say, we know
29:36 you're not all techies and if you bring
29:38 us something that you got your .ru wallet
29:41 and it seems like it was the right idea.
29:43 We want to make you aware of what that
29:44 potential security risk would be. So,
29:46 that's something I expect to see coming
29:48 in the ecosystem.
29:50 Now we at the end of our slide deck
29:53 always go back to what is the
29:56 key points we want and we want you to be
29:59 in charge. We want you to determine who
30:02 you share data with, when you share it
30:04 with them, and why that matters in the
30:07 healthcare environment is we know that
30:10 data can be weaponized against you in
30:12 the health space. And if you
30:15 have no visibility to that and AI models
30:18 are starting to be used and I'll say
30:21 quite frankly in inappropriate ways to
30:22 come to like billing and costing
30:25 decisions, you have to be made aware of
30:27 that. So you have to have control over
30:29 that.
30:32 Interoperability. You've heard this if
30:33 you've been here the last two days.
30:35 What's the interoperability? Well, we
30:38 are going to have to be interoperable.
30:40 We can't expect that every state is
30:42 going to say, you know, we believe
30:44 exactly how Utah believes and we think
30:46 their technology choices are great and
30:49 we are going to go that way.
30:52 I don't know if you noticed during the
30:53 panel with Joe and our representative
30:56 from California, Joe had said we have to
30:59 agree that we're solving the same
31:00 problem. If we're not, then
31:02 interoperability isn't a question. And I
31:05 completely agree with where Joe was at
31:06 on that. We definitely want to make sure
31:08 that we're solving the same issue and
31:10 that we're moving forward in something
31:12 that is rational for all of us. So
31:14 digital reciprocity between states is
31:16 something we're going to have to deal
31:17 with. We definitely want to make sure
31:19 that, hey, if you've ditched all your
31:21 your plastic, you jumped on a plane and
31:23 you're suddenly in Oregon and you get
31:25 pulled over in your rental car, your
31:27 digital driver's license has to be
31:29 something that can be considered valid
31:31 in all jurisdictions. And that's
31:34 quite frankly a format war that it's not
31:36 my problem right now that I'm trying to
31:38 solve. I'm trying to solve the core
31:40 identity issues for Utah residents. Now,
31:42 Of course, we have the
31:45 Olympics coming up 2034.
31:48 There's probably going to be digital
31:49 credentials that come to us from other
31:51 jurisdictions even outside the United
31:53 States. Well, we are building this
31:56 system around the concept of American
31:58 values. And what it is no tracking,
31:59 no surveillance, no, you know, no
32:01 monitoring. So, in those cases, we may
32:03 say, "Yeah, that's a valid credential
32:05 from where you came from." We might
32:06 issue a temporary one while you're here
32:08 for the Olympics to allow you to
32:10 interface and do the things that you
32:11 need to do, but we're certainly not
32:13 going to phone home to a foreign server
32:15 and participate in state surveillance on
32:17 behalf of a of a foreign government. We
32:19 just won't do that. So, when it comes to
32:21 the healthcare interoperability issues,
32:24 we want to make sure that the blast
32:26 radius is always 1. And this is the
32:28 thing that .., I hate to say it, but we're
32:30 finally getting to the point where data
32:32 breaches, people are understanding that
32:34 your data continues to get lost because
32:36 they're in central systems; these huge
32:38 honeypots. If we do this right, people
32:41 are still going to get social
32:42 engineered. But if the blast radius is
32:44 1, that's a much better situation for
32:46 us to be able to handle than having a
32:47 blast radius of, "Oh, 100 million records
32:49 were lost. Would you like some free
32:50 credit monitoring and hope things work
32:53 out for you?" And when we look at the
32:56 overall cyber security risk, there's
32:59 definitely going to be some things that
33:00 we are going to look at from the from
33:01 the wallet provider community, key
33:03 escrow and how that's done is going to
33:05 be super important to us because it's
33:07 not your identity if somebody else has
33:09 full access to your keys and can perform
33:11 rotations without your permission. You
33:13 know, I got asked the other day, well,
33:14 what happens if I'm
33:17 at the border and the law enforcement
33:20 agency requires my phone from me and
33:24 they copy my whole phone and then
33:26 they wander away and now I know they've
33:28 got my current keys. Well, Sam, you can
33:31 put this on a t-shirt if you want, but I
33:32 just said when in doubt, rotate out. You
33:34 don't have to stay with that key pair.
33:36 When you get your phone back, move on to
33:38 your next set. So, I'm going to give you
33:40 a situation that happens in rural
33:42 healthcare because we talked all about
33:43 these things in kind of this really
33:45 large buckets, but when I was down
33:48 talking to rural healthcare, we had to
33:49 have some specific examples of how rural
33:52 healthcare is hurt by not having a good
33:54 identity system that can provide instant
33:57 validation and of assertions of
33:58 credentials. So, if you look up there,
34:01 50,000 to 100,000 is the and this is
34:04 across the nation, not just Utah. this
34:06 peer-reviewed studies that they can
34:08 bring in a doctor or say a critical care
34:10 nurse and getting to the point of having
34:12 the original documentation validated.
34:14 They're basically sweeping floors or
34:16 doing something else or working in a
34:18 non-risk position while those
34:20 credentials get validated. 50 to 100,000
34:23 per provider they bring into the system.
34:25 And the tragedy is when you look at
34:27 these small healthcare systems in these
34:28 rural settings, you may get hospital
34:30 privileges for this rural hospital 50
34:33 miles down the road where you might want
34:35 to work on a weekend or provide
34:37 services. You're not accredited. So you
34:39 have to go through this whole
34:40 credentiing process again. So it's
34:42 really hard for people who have made the
34:44 commitment as a healthcare provider to
34:46 be in a rural community to be able to
34:48 service more than one. It becomes a
34:51 nightmare and a paper castle. So, we're
34:53 looking at SEDI and the subsequent
34:56 credentials that get
34:59 endorsed by the medical
35:01 organizations and others. We want them
35:03 to be able to show up with an ACDC in
35:05 the middle of the night and say, "Here
35:06 you go. Here's my whole chain of
35:07 credentials." And all those boxes get
35:09 ticked and that person goes to work
35:11 doing the job that they want to do in
35:13 that rural community. And of course, one
35:15 of the things that we don't want to
35:18 Underplay is the fact that these
35:20 rural healthcare systems compliance is
35:23 a huge issue. You blow up one thing and
35:25 you suddenly get all your payments
35:26 stopped. In a lot of cases, that can
35:29 be detrimental to the organization. I
35:31 know for a fact that in the state of
35:32 Utah skilled nursing facilities for
35:36 long-term hospice care patients, their
35:39 hash flow is about 2 and a half to 3
35:40 weeks. So, if you get an interruption in
35:42 payments, you're calling relatives and
35:44 saying, "You need to find somewhere else
35:46 because I can't afford to keep my
35:47 facility open." So, it's a real risk.
35:50 All right, let's talk about us. I
35:53 don't know if I'm going to try and get
35:54 this as good as phonehome, but I want to
35:56 kill the clipboard. I am so sick and
36:00 tired of going to an organization that
36:02 has my medical records, and every time I
36:04 come in, they want all this demographic
36:07 data they already have.
36:10 It drives me insane and I hope it drives
36:12 you insane, too, because it's like,
36:13 look, you know who I am. Think if I
36:16 attest to things haven't changed since
36:18 the last time I saw you, but I'm here
36:19 for a sniffle or a broken arm, I don't
36:21 need miles of paperwork to tell you the
36:23 things that I you already know about me.
36:25 And I will say I've been rendered
36:27 functionally illiterate by computers. My
36:29 handwriting is awful. So half the time
36:31 when I'm in these places, they're like,
36:33 "Is that an E or is that an A?" I'm
36:35 like, " I don't know. I'm in
36:37 pain. I'm in a doctor's office. I'm not
36:38 really interested in a critique about my
36:40 my handwriting. And so what we really
36:43 want to do with this kill the clipboard
36:45 is reduce these inaccuracies that come
36:47 in. And for the love of everything, why
36:49 do you have to get a copy a physical
36:51 copy of my medical card when I provided
36:54 it to you a month ago? And part of this
36:56 is not the it's not the fault of the
36:58 front-end worker. They have a set of
37:00 procedures that are trying to do risk
37:02 reduction. And so this is the way that
37:04 we operate our systems. And from my
37:07 perspective, if I could just go in and
37:09 do a tap and say, "Yeah, let me just
37:12 punch the items on the iPad of where I
37:14 hurt and how bad, that's a much better
37:16 experience for me as an individual." And
37:18 I also know that the data errors aren't
37:20 going to be because you couldn't read my
37:21 spelling or you transposed a number on
37:24 my medical card. It all came across in a
37:26 trustable transaction or series of chain
37:28 credentials that we can now move forward
37:30 on the situation. So that's the utopia I
37:33 look for. Now as I said at the SEDI
37:36 summit we look at this in government in
37:39 in a lot of different ways but if you
37:41 take any process whether it's government
37:43 or whether it's business look at the pie
37:46 chart of how much of that transaction is
37:48 trust verification risk management and
37:51 all the other things you actually do
37:52 before you get to service delivery. It's
37:55 inordinate the amount of time we spend
37:57 in making these things work because
37:59 we're we're resetting the trust model
38:01 every single time. In government, where
38:04 this hurts us so much is that we
38:06 actually have to staff people who are
38:08 part of that front office to injust
38:11 this paperwork, make some value
38:13 judgments, move things forward, and it's
38:15 not something that is sustainable. So
38:17 Utah has experienced record growth over
38:20 the last decade. We're just growing,
38:21 growing. Well, we don't want to grow
38:24 government at the same
38:26 rate. I can't keep hiring people to
38:29 ingest this physical paperwork. I have
38:31 to remain in the digital realm, but I
38:32 have to be accurate.
38:36 So, we are facing this dilemma
38:39 everywhere in the United States. As
38:41 populations grow, as you know, civil
38:44 servants retire, next generation may not
38:46 be looking for that sort of work because
38:48 generally it doesn't pay really well. In
38:50 fact, it doesn't ever pay really well.
38:51 But you those are the individuals we
38:54 want to be able to have a good
38:55 experience with us. But we're not just
38:57 going to keep adding and adding and
38:58 adding. So we have to get the digital
39:00 systems working to our best effect.
39:05 So when we look at overall the situation
39:08 that SEDI can solve, it's not a situation
39:11 of where it solves only one problem. It
39:13 solves a number of problems. So when
39:15 you're thinking about your specific
39:18 marketplace that you work in and how
39:20 KERI is going to make things a lot
39:22 better, We see it as an underpinning
39:24 that breaks a lot of the
39:26 traditional models and lets us move on.
39:28 I look forward to a world, you know,
39:30 endpoints where I can identify and
39:32 move things around without an
39:34 intermediary. I know Chris and I have
39:36 talked about this a lot is that when
39:38 this is live and you have these items
39:41 available to you in a functional format
39:43 with utility, we're going to have a lot
39:45 of people who are going, "What's my
39:47 purpose in life now?" Because I've had
39:49 my job as an intermediary for a long
39:51 time. But I'll tell you what, there's
39:53 always going to be something new for
39:54 people to do. I love history and one
39:57 of the things I find funny is when the
39:58 automobiles were first allowed in
40:00 London, it was a requirement that you
40:03 had a signalman walk in front of your
40:05 vehicle with a bell and he'd ring the
40:06 bell to make sure as you put it away
40:09 that the horses were not startled and
40:11 they even extended that into the evening
40:12 where he would carry a lantern. We don't
40:15 have signalmen in front of cars today.
40:17 They do something else. And as we
40:20 move forward with our new user
40:23 centric, you know, sovereign
40:25 environment, I think we're going to see
40:26 some significant changes in
40:28 marketplaces. But it's not just going to
40:30 be driven by the state. We can set
40:31 governance. We can set law. But as a
40:34 larger community, we need to ask those
40:36 questions why we do the things the way
40:37 we do. And if we are not afraid of those
40:40 questions, not only in business, but in
40:42 government, we're going to see these
40:43 massive improvements along the way.
40:45 Okay, we have like three minutes. It's
40:47 always hard to be the one who's like
40:48 backed up against lunch because the
40:50 questions drop off. So, if you don't ask
40:52 your question now, that's okay. Find me
40:54 at lunch or over the next day
40:56 or so. Do we have any questions I can
40:58 answer in the next couple of seconds,
41:01 minutes, whatever.
41:09 Can we get a copy of that?
41:11 Absolutely.
41:14 That was a simple one. Give me a harder
41:17 one.
41:18 - My question is about on
41:21 the delegation slide. You mentioned
41:23 several populations where it's super
41:25 relevant and one of them is like
41:27 unhoused population and migrants and
41:30 what how does delegation look in that
41:33 case? Who's the delegator?
41:37 So in that specific case I'll tell
41:39 you what we find with homeless
41:41 population specifically in the
41:42 healthcare environment. So when we have
41:44 an encounter with somebody who's
41:46 homeless, the best we can do is put them
41:48 on something like emergency Medicaid,
41:49 which is super expensive for the state,
41:51 and we can provide those critical
41:52 services that keep them from
41:54 dying, but that's not long-term
41:57 maintenance. It's not
41:58 prophylactic medications or anything of
42:00 that nature. Those don't come until we
42:02 have identified that individual. And
42:04 given that we might have to query them
42:06 like where were you born approximately
42:08 when were you born all these items that
42:09 we case work it's assigned to them. It
42:12 is generally 6 to 8 weeks before we have
42:15 documentation in hand that we can now
42:17 say we have identified you and we can
42:20 now move you off of the emergency
42:21 Medicaid and put you on to a different
42:23 service. The tragedy is once they
42:26 hand them that packet of information
42:27 within 30 days the majority of them have
42:30 lost their documentation. And it's like
42:32 well of course they're unhoused. they're
42:34 not going to walk around with
42:35 a manilla folder under their arm. So,
42:38 how do you solve that problem? Well,
42:39 there's a couple ways you can do it. One
42:41 of the .., I was at a conference and I
42:43 was just being flip and I said, "Well,
42:45 why don't we do a system called Nomad?"
42:47 And they were like, "What do you mean by
42:49 that?" I said, " National Online
42:52 Assisted Documentation, or
42:54 management of assist of documentation."
42:56 So we are looking at opportunities in it
42:58 that could do something like that which
43:00 is basically there are some things
43:01 you're always going to have with you and
43:03 those are going to be some of your
43:03 biometrics and if you are willing to
43:05 work with the state to get you
43:07 identified and put you into a national
43:10 data store that's encrypted to your
43:12 biometrics then we have means for how we
43:14 can get those people on the system.
43:16 There are care relationships that exist
43:18 inside the state like Foster Care or we
43:20 have Battered Women who are put in like
43:23 safe harbor shelters. So there are some
43:26 there are some relationships that we
43:28 need to clarify in not only in law but
43:31 how that works digitally because there
43:32 are certain rights that get communicated
43:34 to each group. So as we move along there
43:36 will be both a policy side and a
43:38 software technology side to work on it.
43:40 But the interesting thing is we do
43:43 think that the easiest relationship for
43:45 us to solve right off the bat is the
43:47 parent child relationship. That's pretty
43:49 easy to prove. But when we get into some
43:51 of these more complicated
43:52 situations, so say for example, a
43:55 Medicaid household is defined by the
43:57 number of people living under the same
43:58 roof and what happens if you have an
44:01 acrimonious divorce. Party A has a lot
44:04 of KYC data to try and infiltrate what
44:07 party B is doing. So the state's going
44:08 to have to be very careful as we roll
44:10 these out to make sure that when we do
44:11 endorse something like a parent child
44:13 relationship that we do have a glass
44:15 break where we can go and turn that off
44:17 so that we keep everyone safe.
44:20 Okay, I think you is we're at lunch.
44:22 So, thank you again. I look forward to
44:23 talking to you individually. Reach me
44:25 out to me at any time. I'm always
44:27 willing to talk about both the tech and
44:29 the policy. Thank you very much.
44:32 Thanks.