1 00:00:00,000 --> 00:00:02,240 Before we jump into the source material today 2 00:00:02,240 --> 00:00:06,910 We really need to address this fundamental problem that honestly almost every 3 00:00:06,910 --> 00:00:09,220 organization rustles with at some point 4 00:00:09,220 --> 00:00:14,420 Yeah, the sheer cost and well the incredible complexity of secure logins, right? 5 00:00:14,420 --> 00:00:15,040 Exactly 6 00:00:15,040 --> 00:00:19,660 And that brings us to the supporter of today's deep dive safe server 7 00:00:19,660 --> 00:00:24,130 Yeah, if you're building an app or you know managing an organization's digital 8 00:00:24,130 --> 00:00:25,200 infrastructure right now 9 00:00:25,260 --> 00:00:29,790 You are probably looking at these incredibly expensive proprietary authentication 10 00:00:29,790 --> 00:00:32,440 tools. Oh, yeah things like author or 11 00:00:32,440 --> 00:00:37,820 Microsoft entra Google Cloud identity is another big one, right and safe server 12 00:00:37,820 --> 00:00:40,900 basically helps you implement really powerful open source 13 00:00:40,900 --> 00:00:45,760 Alternatives to those massive, you know proprietary systems and why make the switch? 14 00:00:45,760 --> 00:00:48,620 I mean beyond the massive cost savings, which is huge 15 00:00:48,620 --> 00:00:52,300 It really comes down to data sovereignty, which is so critical these days 16 00:00:52,300 --> 00:00:55,940 Exactly when you are dealing with legal and compliance requirements like strict 17 00:00:55,940 --> 00:00:57,540 data protection email retention 18 00:00:57,540 --> 00:01:02,550 Financial records and audit trails you really cannot afford to have your users 19 00:01:02,550 --> 00:01:03,340 identity data 20 00:01:03,340 --> 00:01:08,420 Just locked away in a vendor's black box. You need to own it, right and safe server 21 00:01:08,420 --> 00:01:10,840 guys organizations through that entire process 22 00:01:10,840 --> 00:01:14,430 Yeah from you know that initial consulting phase to figure out exactly what 23 00:01:14,430 --> 00:01:16,420 architecture you actually need all the way to 24 00:01:16,800 --> 00:01:21,270 Operating the open source software on highly secure servers located right within 25 00:01:21,270 --> 00:01:24,140 the EU which solves a ton of headaches for compliance 26 00:01:24,140 --> 00:01:28,610 Huge headaches you keep control you stay compliant and you stop paying that 27 00:01:28,610 --> 00:01:30,180 ridiculous enterprise markup 28 00:01:30,180 --> 00:01:33,740 You can find more information and get started at 29 00:01:33,740 --> 00:01:36,140 www.safeserver.de 30 00:01:36,140 --> 00:01:38,060 Because honestly 31 00:01:38,060 --> 00:01:42,360 Outsourcing your core identity infrastructure to those proprietary giants 32 00:01:42,360 --> 00:01:46,920 It's just a massive strategic risk and we are seeing a really significant industry 33 00:01:46,920 --> 00:01:48,840 pivot toward reclaiming that control 34 00:01:48,840 --> 00:01:51,520 Yeah, and speaking of risks if you're building an app today 35 00:01:51,520 --> 00:01:56,260 You might have this like massive blind spot as you're probably focusing entirely on 36 00:01:56,260 --> 00:01:57,860 human users, right? 37 00:01:57,860 --> 00:02:01,880 People clicking buttons exactly but looking at the trajectory of modern networks 38 00:02:01,880 --> 00:02:07,290 Non-human AI agents are well, they're likely going to outnumber human users in the 39 00:02:07,290 --> 00:02:08,540 very near future 40 00:02:08,540 --> 00:02:11,420 It's not even a question of if but when so true 41 00:02:11,780 --> 00:02:16,600 So today we're diving into this whole stack of github repositories security white 42 00:02:16,600 --> 00:02:18,100 papers and developer docs 43 00:02:18,100 --> 00:02:21,440 centered around an open source authentication and 44 00:02:21,440 --> 00:02:26,860 Authorization infrastructure called log toe log toe and our goal today is to really 45 00:02:26,860 --> 00:02:29,340 analyze how modern applications are 46 00:02:29,340 --> 00:02:35,220 Attempting to secure themselves for both human users and these you know AI agents 47 00:02:35,220 --> 00:02:40,340 Yeah, and how they do it without forcing developers to become like full-blown cryptographers 48 00:02:40,340 --> 00:02:41,680 because authentication remains 49 00:02:42,440 --> 00:02:45,990 The absolute bane of every software developers existence. I mean, it's just a 50 00:02:45,990 --> 00:02:48,880 nightmare. It really is. Why is that though? 51 00:02:48,880 --> 00:02:53,350 Well, the friction really comes from the gap between user expectation and technical 52 00:02:53,350 --> 00:02:53,800 reality 53 00:02:53,800 --> 00:02:57,620 So a user opens an app they click a button and they just expect to be in right 54 00:02:57,620 --> 00:02:58,960 compose that right? 55 00:02:58,960 --> 00:03:03,520 But beneath that single click is this terrifying labyrinth of security protocols 56 00:03:03,520 --> 00:03:04,120 Yeah 57 00:03:04,120 --> 00:03:07,680 I want to look at the actual mechanics of that labyrinth because the documentation 58 00:03:07,680 --> 00:03:10,040 we're looking at frequently cites these protocols 59 00:03:10,400 --> 00:03:12,920 like OIDC, which is open ID connect and 60 00:03:12,920 --> 00:03:16,320 Always 2.1. It's a big ones. Yeah, right 61 00:03:16,320 --> 00:03:21,030 But why are these so painfully difficult for a standard product developer to just 62 00:03:21,030 --> 00:03:22,480 implement from scratch? 63 00:03:22,480 --> 00:03:25,960 I mean building an app and then having to build your own authentication system 64 00:03:25,960 --> 00:03:28,400 It feels like I don't know like opening a local bakery 65 00:03:28,400 --> 00:03:32,150 But then having to design and forge your own bank vault from scratch just to hold 66 00:03:32,150 --> 00:03:33,280 the cash register 67 00:03:33,280 --> 00:03:37,420 That is actually a perfect analogy because it comes down to the catastrophic cost 68 00:03:37,420 --> 00:03:38,360 of a tiny mistake 69 00:03:38,520 --> 00:03:43,450 You know OIDC handles the authentication part that basically verifying who the user 70 00:03:43,450 --> 00:03:44,080 actually is 71 00:03:44,080 --> 00:03:47,580 Okay, and then oh, I have two point one handles the authorization 72 00:03:47,580 --> 00:03:53,130 So granting permissions to access certain data but without handing over the actual 73 00:03:53,130 --> 00:03:54,040 password, right? 74 00:03:54,040 --> 00:03:56,830 You don't want to just pass passwords around exactly. So when you use these 75 00:03:56,830 --> 00:03:59,680 protocols, you aren't just sending a username back and forth 76 00:03:59,680 --> 00:04:06,860 You are generating a cryptographically signed JSON web token or a JWT 77 00:04:06,860 --> 00:04:11,000 Okay JWT and what happens if the developer, you know 78 00:04:11,000 --> 00:04:15,220 Just writes the logic for that token incorrectly like a small typo or something 79 00:04:15,220 --> 00:04:17,120 total compromise like absolute disaster 80 00:04:17,120 --> 00:04:19,120 We really just from a small error. Oh, yeah 81 00:04:19,120 --> 00:04:25,190 Let's say a developer accidentally configures their system to accept a token signed 82 00:04:25,190 --> 00:04:25,960 with the wrong 83 00:04:25,960 --> 00:04:30,280 Cryptographic algorithm like they confuse a symmetric key for an asymmetric one 84 00:04:30,280 --> 00:04:35,300 Oh, wow, right a malicious actor can then forge their own digital pass 85 00:04:35,960 --> 00:04:38,100 Rewrite the payload inside it to say hey 86 00:04:38,100 --> 00:04:42,350 I'm the administrator and the system will just let them write in that is terrifying. 87 00:04:42,350 --> 00:04:46,240 It is implementing these industry standard protocols from scratch 88 00:04:46,240 --> 00:04:51,960 Requires like a really deep understanding of token life cycles key rotation 89 00:04:51,960 --> 00:04:56,970 Cryptographic validation. It's not for beginners. So looking at the developer docs 90 00:04:56,970 --> 00:05:00,100 here log toes primary value proposition seems to be just 91 00:05:00,100 --> 00:05:04,480 You know removing that cryptographic burden entirely going back to the analogy 92 00:05:04,480 --> 00:05:09,420 It acts as that pre-forged highly secure bank vault exactly developer just drops it 93 00:05:09,420 --> 00:05:11,160 into their back end and log toe handles 94 00:05:11,160 --> 00:05:15,060 All the OIDC and o with 2.1 token verification 95 00:05:15,060 --> 00:05:19,280 Natively that is the core utility right there the developer doesn't have to write 96 00:05:19,280 --> 00:05:21,320 the intricate logic to validate a web token 97 00:05:21,320 --> 00:05:24,270 They can just focus on building their actual product rather than you know trying to 98 00:05:24,270 --> 00:05:26,320 become an identity security expert overnight 99 00:05:26,320 --> 00:05:31,020 Which is a huge relief right and from a front-end perspective the sources really 100 00:05:31,020 --> 00:05:33,900 detail how this translates into the user experience, right? 101 00:05:33,900 --> 00:05:38,530 Yeah, the visual side of it right log to provides these pre-built sign-in flows. So 102 00:05:38,530 --> 00:05:40,640 you get passwordless entry via email or 103 00:05:40,640 --> 00:05:47,660 SMS verification codes you have traditional passwords obviously and social sign-ins 104 00:05:47,660 --> 00:05:50,720 like Google Apple or discord the ones everyone expects 105 00:05:50,720 --> 00:05:54,850 Yeah, but the feature that really stands out to me in the documentation is this 106 00:05:54,850 --> 00:05:57,240 multi app omni sign-in experience 107 00:05:57,240 --> 00:06:02,420 Oh, that is huge. That solves a major psychological friction point for users. How 108 00:06:02,420 --> 00:06:03,700 so well imagine an 109 00:06:03,900 --> 00:06:08,010 Organization that has like a web dashboard a mobile app and maybe the community 110 00:06:08,010 --> 00:06:08,280 forum 111 00:06:08,280 --> 00:06:12,600 Normally a user might navigate between those platforms and encounter three 112 00:06:12,600 --> 00:06:14,760 completely different login screens 113 00:06:14,760 --> 00:06:19,790 Oh, I hate that right because that inconsistency breeds suspicion users wonder if 114 00:06:19,790 --> 00:06:22,280 they've stumbled onto a phishing site or something 115 00:06:22,280 --> 00:06:23,800 Oh, okay. Yeah, that makes total sense 116 00:06:23,800 --> 00:06:24,000 Yeah 117 00:06:24,000 --> 00:06:28,500 So with the omni sign-in approach the login screen just looks native and consistent 118 00:06:28,500 --> 00:06:30,960 no matter where the user encounters it exactly 119 00:06:30,960 --> 00:06:35,920 It's one centralized sign-in flow for every single application in that ecosystem 120 00:06:35,920 --> 00:06:40,020 Okay, but I want to push back on this omni sign-in idea for a second just based on 121 00:06:40,020 --> 00:06:42,280 the enterprise use cases mentioned in the sources 122 00:06:42,280 --> 00:06:46,820 Okay, let's hear it having one unified login screen sounds really great for the 123 00:06:46,820 --> 00:06:48,320 software provider, right? 124 00:06:48,320 --> 00:06:53,410 Yeah, but what if that software provider grows and lands this massive corporate 125 00:06:53,410 --> 00:06:53,940 client? 126 00:06:53,940 --> 00:06:59,310 Ah the b2b scaling problem, right if I am a large corporation renting space on this 127 00:06:59,310 --> 00:06:59,960 platform 128 00:06:59,960 --> 00:07:04,970 I do not want my employees seeing a generic log to a screen. I want my own 129 00:07:04,970 --> 00:07:06,200 corporate branding 130 00:07:06,200 --> 00:07:11,000 There's a system like this force a universal look or does it you know allow for? 131 00:07:11,000 --> 00:07:15,600 Customization at the client level that is a critical distinction to make and the 132 00:07:15,600 --> 00:07:18,760 documentation actually specifically addresses this through a concept called 133 00:07:18,760 --> 00:07:23,400 Multi-tenancy multi-tenancy. Yeah log tow does not force a universal look on 134 00:07:23,400 --> 00:07:24,520 corporate clients 135 00:07:24,520 --> 00:07:29,080 It is built from the ground up to support business to business or b2b scaling 136 00:07:29,080 --> 00:07:30,880 Okay, let's unpack the mechanics of that 137 00:07:30,880 --> 00:07:35,010 How does multi-tenancy actually isolate one client from another think of the 138 00:07:35,010 --> 00:07:36,420 application as a large? 139 00:07:36,420 --> 00:07:41,930 Apartment building when you have individual consumer users. They each basically get 140 00:07:41,930 --> 00:07:43,560 a small isolated apartment 141 00:07:43,560 --> 00:07:48,120 Okay makes sense, but when you land a corporate client, let's call them business a 142 00:07:48,120 --> 00:07:50,120 they don't just want one apartment 143 00:07:50,120 --> 00:07:52,040 They want to rent an entire floor, right? 144 00:07:52,040 --> 00:07:56,150 They have a whole team exactly and they want the master keys to manage all the 145 00:07:56,150 --> 00:07:57,680 individual rooms on that floor 146 00:07:58,240 --> 00:08:03,660 So log toe allows developers to create distinct organizations or tenants within the 147 00:08:03,660 --> 00:08:03,960 app 148 00:08:03,960 --> 00:08:08,770 Business a gets its own isolated tenant environment. So they get their own branding 149 00:08:08,770 --> 00:08:09,840 their own user grouping 150 00:08:09,840 --> 00:08:14,310 Yep, and most importantly strict data isolation from business B on the floor right 151 00:08:14,310 --> 00:08:15,360 below them. Got it 152 00:08:15,360 --> 00:08:22,190 So business a is database queries cannot accidentally like bleed over and expose 153 00:08:22,190 --> 00:08:23,640 business B's user data 154 00:08:24,080 --> 00:08:28,370 Precisely, that would be a massive compliance failure and within that isolated 155 00:08:28,370 --> 00:08:31,900 floor. The corporate client also needs our BAC our BAC 156 00:08:31,900 --> 00:08:35,150 Which is role-based access control, right? Right? And this is another one of those 157 00:08:35,150 --> 00:08:36,200 enterprise grade features 158 00:08:36,200 --> 00:08:38,560 That's just native to lock toe. So under the hood 159 00:08:38,560 --> 00:08:43,250 How does our BAC actually restrict someone is it literally just a list of names on 160 00:08:43,250 --> 00:08:44,300 a server? No 161 00:08:44,300 --> 00:08:46,560 It's much more robust than that our BAC 162 00:08:46,560 --> 00:08:51,130 Attaches specific permissions to roles and then assigns those roles to users via 163 00:08:51,130 --> 00:08:53,800 claims in their actual identity token 164 00:08:53,800 --> 00:08:56,320 Okay, so like a tag on their ID badge basically 165 00:08:56,320 --> 00:09:01,330 So an intern receives a token with say a read-only claim while a manager receives a 166 00:09:01,330 --> 00:09:03,280 token with a write and delete claim 167 00:09:03,280 --> 00:09:08,040 When that intern tries to delete a database the application checks the token 168 00:09:08,040 --> 00:09:12,510 Sees the missing permission and just blocks the action at the protocol level. That's 169 00:09:12,510 --> 00:09:13,080 incredibly secure 170 00:09:13,080 --> 00:09:15,960 It is and lock toe manages all of this logic 171 00:09:15,960 --> 00:09:20,070 Allowing the corporate client to set strict rules without the software developer 172 00:09:20,070 --> 00:09:21,240 having to manually code 173 00:09:21,360 --> 00:09:25,560 You know permission checks for every single user Wow and the sources also highlight 174 00:09:25,560 --> 00:09:27,000 enterprise SSO 175 00:09:27,000 --> 00:09:31,660 So single sign-on as the sort of holy grail of b2b software. We are talking about 176 00:09:31,660 --> 00:09:32,360 integrations with 177 00:09:32,360 --> 00:09:34,080 semel 178 00:09:34,080 --> 00:09:36,080 Entra or Okta 179 00:09:36,080 --> 00:09:40,760 Why do large corporations mandate this so strictly it is entirely about? 180 00:09:40,760 --> 00:09:45,230 Compliance and the IT department's ability to maintain a single kill switch a kill 181 00:09:45,230 --> 00:09:45,400 switch 182 00:09:45,400 --> 00:09:51,640 Yeah, if a corporation has like 5,000 employees the IT department cannot manage 183 00:09:51,640 --> 00:09:56,880 5,000 separate accounts across 50 different software platforms. It's impossible. 184 00:09:56,880 --> 00:09:57,800 Yeah, that would be a nightmare 185 00:09:57,800 --> 00:10:02,810 So they want one centralized identity provider like Okta when an employee leaves 186 00:10:02,810 --> 00:10:06,200 the company IT just disables their Okta account 187 00:10:06,200 --> 00:10:10,720 Once oh and that instantly revokes their access to every single app they used 188 00:10:10,720 --> 00:10:11,440 including yours 189 00:10:11,720 --> 00:10:16,190 Exactly log toe allows developers to plug directly into these enterprise identity 190 00:10:16,190 --> 00:10:16,720 providers 191 00:10:16,720 --> 00:10:20,850 Satisfying those corporate compliance requirements right out of the box. Okay, but 192 00:10:20,850 --> 00:10:21,280 normally 193 00:10:21,280 --> 00:10:26,050 Providing features like SAML integration and multi-tenancy that requires enterprise 194 00:10:26,050 --> 00:10:26,960 level pricing 195 00:10:26,960 --> 00:10:31,730 But analyzing log toes pricing model here reveals a really interesting strategy the 196 00:10:31,730 --> 00:10:32,080 first 197 00:10:32,080 --> 00:10:36,600 50,000 monthly active users are completely free which is wild 198 00:10:36,600 --> 00:10:40,300 It functions as this massive safety net for growing startup by eliminating the 199 00:10:40,300 --> 00:10:42,480 upfront cost for those enterprise features 200 00:10:42,480 --> 00:10:47,040 Developers can architect their platform for B2B scale from day one 201 00:10:47,040 --> 00:10:51,510 Rather than you know trying to duct tape multi-tenancy onto a consumer app three 202 00:10:51,510 --> 00:10:53,840 years down the line when they finally land a big contract 203 00:10:53,840 --> 00:10:57,970 Yeah, buck taping enterprise features is never a good idea never now the 204 00:10:57,970 --> 00:10:59,920 documentation makes a really sharp pivot here 205 00:10:59,920 --> 00:11:04,800 Moving away from human users entirely and this brings us back to that blind spot 206 00:11:04,800 --> 00:11:07,640 We mentioned at the start of the show the AI agents. Yes 207 00:11:07,640 --> 00:11:12,480 How does an authentication system handle the paradigm shift of the AI era? 208 00:11:12,480 --> 00:11:18,200 I mean we're scripts microservices and AI models need to securely access data 209 00:11:18,200 --> 00:11:22,400 Because if our BAC is like giving different employees different colored key cards 210 00:11:22,400 --> 00:11:26,600 How do we give a key card to a robot assistant who needs to do tasks on our behalf? 211 00:11:26,600 --> 00:11:30,920 This is where identity infrastructure is currently undergoing a massive evolution 212 00:11:31,440 --> 00:11:35,910 historically, you know authentication required a screen a keyboard and human eyes 213 00:11:35,910 --> 00:11:37,760 to solve a KPT CHA or 214 00:11:37,760 --> 00:11:41,880 Read an SMS code right click all the squares of the crosswalk 215 00:11:41,880 --> 00:11:45,640 Exactly, but AI agents do not have eyes or smartphones 216 00:11:45,640 --> 00:11:50,190 So Logto is engineering identity for these non-human entities through machine to 217 00:11:50,190 --> 00:11:53,440 machine or M2M authentication 218 00:11:53,440 --> 00:11:55,440 And the documentation 219 00:11:55,440 --> 00:12:00,320 Mentions the model context protocol or MCP in this specific context. What role does 220 00:12:00,320 --> 00:12:00,720 that play? 221 00:12:00,720 --> 00:12:06,720 So MCP is this emerging standard designed to securely connect AI models to external 222 00:12:06,720 --> 00:12:08,840 tools and private data sources 223 00:12:08,840 --> 00:12:09,240 Right 224 00:12:09,240 --> 00:12:14,000 If you deploy an AI assistant and ask it to summarize say a highly confidential 225 00:12:14,000 --> 00:12:16,120 financial report from your private database 226 00:12:16,120 --> 00:12:20,760 That AI needs a way to cryptographically prove to the database that it actually has 227 00:12:20,760 --> 00:12:23,880 the authority to read that file and Logto handles that 228 00:12:23,880 --> 00:12:28,350 Right Logto natively supports the protocols required for those secure automated 229 00:12:28,350 --> 00:12:30,560 handshakes and the mechanism they use for this is called 230 00:12:30,720 --> 00:12:32,720 Personal access tokens or BATs 231 00:12:32,720 --> 00:12:35,640 How do these differ from a standard login token? 232 00:12:35,640 --> 00:12:39,210 Think of it this way a standard login is like presenting your passport at a border 233 00:12:39,210 --> 00:12:40,200 crossing, right? 234 00:12:40,200 --> 00:12:44,320 The guard has to look at your face to verify. It's really you a personal access 235 00:12:44,320 --> 00:12:45,800 token is more like a diplomatic pouch 236 00:12:45,800 --> 00:12:51,610 It skips the face check entirely, but it is strictly limited in scope. You generate 237 00:12:51,610 --> 00:12:54,040 a PT specifically for your automated pipeline 238 00:12:54,040 --> 00:12:59,320 It provides clean programmatic access without any messy login screens 239 00:12:59,320 --> 00:13:03,590 And I assume the major security benefit there is revocation like if the script goes 240 00:13:03,590 --> 00:13:05,360 rogue or the token is leaked somehow 241 00:13:05,360 --> 00:13:08,160 You don't have to change a master password for the whole system 242 00:13:08,160 --> 00:13:12,820 Exactly. You just instantly revoke that specific diplomatic pouch and the AI agent 243 00:13:12,820 --> 00:13:14,400 loses access immediately 244 00:13:14,400 --> 00:13:19,220 That is the exact security posture you need for modern micro services. It isolates 245 00:13:19,220 --> 00:13:20,480 the risk completely 246 00:13:20,480 --> 00:13:24,000 There is another feature detailed in the sources that sort of bridges this gap 247 00:13:24,000 --> 00:13:27,840 between human support and secure access and it's called 248 00:13:28,160 --> 00:13:29,600 impersonation the 249 00:13:29,600 --> 00:13:33,710 Documentation describes this as a way for customer support teams to log in as an 250 00:13:33,710 --> 00:13:34,440 end customer 251 00:13:34,440 --> 00:13:38,550 Yeah, this is a tricky one right because initially that sounds like a massive 252 00:13:38,550 --> 00:13:39,800 security vulnerability 253 00:13:39,800 --> 00:13:44,730 How does a system allow an employee to inhabit a user's account without you know 254 00:13:44,730 --> 00:13:46,320 exposing the users password? 255 00:13:46,320 --> 00:13:50,840 It uses an audit logged temporary session token. Okay, break that down for me 256 00:13:50,840 --> 00:13:55,310 So when a user reports a broken dashboard the support representative doesn't ask 257 00:13:55,310 --> 00:13:56,200 for their password 258 00:13:56,200 --> 00:13:58,200 They never see the password 259 00:13:58,200 --> 00:14:03,720 Instead the rep uses their own high-level admin credentials to request an impersonation 260 00:14:03,720 --> 00:14:05,360 token from Logto 261 00:14:05,360 --> 00:14:10,550 Okay, and then the system generates a temporary highly restricted session that 262 00:14:10,550 --> 00:14:13,560 basically mirrors the users view and crucially 263 00:14:13,560 --> 00:14:18,390 Every single action the support rep takes while impersonating that user is 264 00:14:18,390 --> 00:14:20,280 permanently logged in the audit trail 265 00:14:20,280 --> 00:14:24,410 Oh under the support reps name not the users exactly it provides the visibility 266 00:14:24,410 --> 00:14:25,440 needed for troubleshooting 267 00:14:25,720 --> 00:14:30,240 While maintaining perfect cryptographic accountability that is brilliant 268 00:14:30,240 --> 00:14:33,880 Okay, so we're looking at an impressive list of capabilities here multi-tenancy 269 00:14:33,880 --> 00:14:37,160 machine-to-machine tokens support impersonation 270 00:14:37,160 --> 00:14:41,970 But the practical reality for developer is you know implementation right building 271 00:14:41,970 --> 00:14:42,360 the thing 272 00:14:42,360 --> 00:14:42,600 Yeah 273 00:14:42,600 --> 00:14:44,520 if Logto is open source 274 00:14:44,520 --> 00:14:49,230 Does a developer just download a massive repository of complex code from github and 275 00:14:49,230 --> 00:14:51,800 just hope they can compile it without breaking their servers 276 00:14:51,840 --> 00:14:56,040 Well, the open-source ecosystem has thankfully evolved significantly past that 277 00:14:56,040 --> 00:14:56,360 point 278 00:14:56,360 --> 00:15:00,960 Logto has really mitigated the integration friction by providing these pre-built 279 00:15:00,960 --> 00:15:03,520 software development kits or SDKs 280 00:15:03,520 --> 00:15:07,120 For over 30 different frameworks 30 Wow 281 00:15:07,120 --> 00:15:07,800 Yeah 282 00:15:07,800 --> 00:15:12,050 So whether a team is building the front end in reactor view and the back end in 283 00:15:12,050 --> 00:15:14,900 Python go iOS Android node 284 00:15:14,900 --> 00:15:19,920 There's a native SDK designed to just handle the API calls to Logto 285 00:15:20,160 --> 00:15:23,660 Effortlessly and the developer experience detailed in the launch options is also 286 00:15:23,660 --> 00:15:24,500 really revealing 287 00:15:24,500 --> 00:15:29,520 They highlight a git pod launch option alongside a local Docker compose setup for 288 00:15:29,520 --> 00:15:30,900 someone new to this 289 00:15:30,900 --> 00:15:35,200 Why would a developer choose one over the other it really comes down to environment 290 00:15:35,200 --> 00:15:37,760 variables and speed a local Docker? 291 00:15:37,760 --> 00:15:41,390 Environment requires you to actually download the containers and run them on your 292 00:15:41,390 --> 00:15:43,520 own machine, which takes time, right? 293 00:15:43,520 --> 00:15:45,000 It is highly reproducible 294 00:15:45,000 --> 00:15:48,790 But it relies on your local hardware git pod on the other hand spins up the entire 295 00:15:48,790 --> 00:15:51,320 open source environment in the cloud instantly 296 00:15:51,320 --> 00:15:55,790 Oh, so you don't install anything locally. No a developer can click a link 297 00:15:55,790 --> 00:15:59,060 completely bypass all local configuration issues and 298 00:15:59,060 --> 00:16:03,170 Just start testing the authentication flows in their browser within seconds that 299 00:16:03,170 --> 00:16:05,160 speed of deployment is impressive, but 300 00:16:05,160 --> 00:16:11,160 We have to analyze the actual defensive architecture here if we return to that bank 301 00:16:11,160 --> 00:16:11,860 vault analogy 302 00:16:11,860 --> 00:16:16,120 We really need to know how thick the walls are absolutely because open source code 303 00:16:16,120 --> 00:16:18,060 is inherently transparent, right? 304 00:16:18,060 --> 00:16:21,100 Which means hackers can study it just as easily as developers 305 00:16:21,100 --> 00:16:24,810 Well transparency is actually a defensive advantage in cryptography because it 306 00:16:24,810 --> 00:16:25,320 allows 307 00:16:25,320 --> 00:16:30,150 Thousands of independent security researchers to audit the code. They find the bugs 308 00:16:30,150 --> 00:16:32,200 before the bad guys do that's true 309 00:16:32,200 --> 00:16:37,240 But beyond the code itself log toe undergoes rigorous institutional auditing 310 00:16:37,280 --> 00:16:42,880 They are soc2 type 2 certified. Okay. What does that certification actually entail? 311 00:16:42,880 --> 00:16:46,670 It's just like a security checklist a company fills out once a year and says we're 312 00:16:46,670 --> 00:16:46,960 good 313 00:16:46,960 --> 00:16:51,230 Not at all. So c2 type I is a checklist that proves you have security policies in 314 00:16:51,230 --> 00:16:52,240 place on a specific day 315 00:16:52,240 --> 00:16:56,080 so c2 type 2 is a grueling continuous audit an 316 00:16:56,080 --> 00:17:00,780 Independent auditor basically monitors the company's operational practices over an 317 00:17:00,780 --> 00:17:03,560 extended period often six months to a year 318 00:17:03,560 --> 00:17:07,670 Wow, just watching everything they do everything to prove that they actually follow 319 00:17:07,670 --> 00:17:12,120 their strict security privacy and data access procedures every single day 320 00:17:12,120 --> 00:17:17,300 Achieving it is incredibly difficult for a startup to do alone. Okay, that gives a 321 00:17:17,300 --> 00:17:17,880 lot of confidence 322 00:17:17,880 --> 00:17:23,010 Yeah, and delving into the actual cryptography the white papers specify that log to 323 00:17:23,010 --> 00:17:25,720 uses argon 2 for password hashing 324 00:17:25,720 --> 00:17:30,000 Yep, argon 2. How does argon 2 protect a database if it actually gets stolen? 325 00:17:30,160 --> 00:17:34,440 Well when a secure system saves a password it never saves the plain text obviously 326 00:17:34,440 --> 00:17:37,660 it runs the password through a mathematical hash function to create 327 00:17:37,660 --> 00:17:42,130 Scramble text right if a hacker steals the database. They usually deploy massive 328 00:17:42,130 --> 00:17:44,840 arrays of graphics processing units GPUs to guess 329 00:17:44,840 --> 00:17:49,120 Millions of passwords a second just hoping to find the text that matches the hash 330 00:17:49,120 --> 00:17:49,920 root force 331 00:17:49,920 --> 00:17:54,080 Exactly. Yeah, argon 2 is designed to defeat this specific hardware attack 332 00:17:54,080 --> 00:17:57,640 How does it bottle like a GPU though like GPUs are incredibly fast 333 00:17:57,680 --> 00:18:02,000 They are but argon 2 defeats them by utilizing memory hard functions 334 00:18:02,000 --> 00:18:08,060 See GPUs have thousands of processing cores allowing them to calculate incredibly 335 00:18:08,060 --> 00:18:10,740 fast, but they have very little local memory 336 00:18:10,740 --> 00:18:16,560 Right argon 2 forces the hashing process to consume a massive amount of RAM 337 00:18:16,560 --> 00:18:22,260 By demanding memory rather than just processing speed argon 2 starves the GPU 338 00:18:22,840 --> 00:18:27,880 Slowing the hackers brute-force attack to an absolute crawl. That is so clever 339 00:18:27,880 --> 00:18:31,200 It essentially turns the attackers own hardware architecture against them 340 00:18:31,200 --> 00:18:35,670 Exactly and for organizations that want to eliminate passwords entirely the 341 00:18:35,670 --> 00:18:38,240 documentation also highlights support for a web often 342 00:18:38,240 --> 00:18:41,520 Which is the underlying technology for paskies right paskies are everywhere now 343 00:18:41,520 --> 00:18:44,960 because web often represents the gold standard for multi-factor authentication 344 00:18:44,960 --> 00:18:49,430 Instead of relying on a shared secret like a password it uses public key cryptography 345 00:18:49,430 --> 00:18:52,280 tied to the biometric sensors on your actual device 346 00:18:52,280 --> 00:18:57,290 Okay, wait, so when I use my fingerprint or face ID to log in my biometric data isn't 347 00:18:57,290 --> 00:18:58,640 being sent to log toe server 348 00:18:58,640 --> 00:19:03,350 Correct. Your biometric data never ever leaves the secure on play of your phone or 349 00:19:03,350 --> 00:19:05,340 laptop. Oh, thank goodness. Yeah 350 00:19:05,340 --> 00:19:11,420 Your device uses that biometric check locally to unlock a private cryptographic key 351 00:19:11,420 --> 00:19:16,130 Which then signs a challenge sent by the server the server only holds the public 352 00:19:16,130 --> 00:19:17,840 key which is useless to a hacker 353 00:19:18,080 --> 00:19:21,730 So even if the server is compromised, they don't have your face or fingerprint 354 00:19:21,730 --> 00:19:25,660 exactly it fundamentally eliminates the risk of phishing and credential stuffing 355 00:19:25,660 --> 00:19:30,130 The final aspect of the architecture detailed in the sources really brings us back 356 00:19:30,130 --> 00:19:31,420 to the deployment models 357 00:19:31,420 --> 00:19:34,800 While log to is open source and can be self hosted 358 00:19:34,800 --> 00:19:39,390 They also offer a fully managed log to cloud version right for teams that don't 359 00:19:39,390 --> 00:19:41,800 want to manage servers and this highlights a critical 360 00:19:41,800 --> 00:19:46,420 Geographical feature regarding data isolation across distinct regions, right? 361 00:19:46,420 --> 00:19:50,550 Yes, and this is a vital component of international compliance log to cloud 362 00:19:50,550 --> 00:19:53,160 operates environments in the EU the US 363 00:19:53,160 --> 00:19:54,940 Australia and Japan 364 00:19:54,940 --> 00:19:59,070 Meaning a European company can ensure their users identity data literally never 365 00:19:59,070 --> 00:20:02,240 crosses the Atlantic never it stays in Europe 366 00:20:02,240 --> 00:20:06,350 Fully encrypted at the database level with TLS encryption securing the data while 367 00:20:06,350 --> 00:20:07,160 it is in transit 368 00:20:07,160 --> 00:20:12,320 It allows an organization to really leverage the transparency and flexibility of an 369 00:20:12,320 --> 00:20:13,560 open source architecture 370 00:20:14,160 --> 00:20:18,290 While still meeting the rigorous data sovereignty demands of an enterprise 371 00:20:18,290 --> 00:20:19,920 environment exactly 372 00:20:19,920 --> 00:20:23,640 However, it is important to note that maintaining self-hosted identity 373 00:20:23,640 --> 00:20:25,840 infrastructure does require 374 00:20:25,840 --> 00:20:30,580 Dedicated engineering resources to manage updates and security patches right? You 375 00:20:30,580 --> 00:20:31,540 can't just set it and forget it 376 00:20:31,540 --> 00:20:36,320 No log toe removes the cryptographic complexity, but the infrastructure still 377 00:20:36,320 --> 00:20:38,040 demands operational diligence 378 00:20:38,720 --> 00:20:42,400 Synthesizing the source material here log toe presents a genuinely compelling 379 00:20:42,400 --> 00:20:44,400 approach to modern identity management 380 00:20:44,400 --> 00:20:49,040 It essentially democratizes access to enterprise-grade security tools 381 00:20:49,040 --> 00:20:53,970 It really does developers are handed this pre-built cryptographic vault that scales 382 00:20:53,970 --> 00:20:54,680 from you know 383 00:20:54,680 --> 00:20:58,950 A simple consumer app all the way to a massive multi-tenant b2b platform with 384 00:20:58,950 --> 00:21:01,320 single sign on and most importantly 385 00:21:01,320 --> 00:21:06,080 It bridges the gap into the AI frontier providing the machine to machine protocols 386 00:21:06,080 --> 00:21:08,360 necessary to secure non-human agents 387 00:21:08,360 --> 00:21:12,160 Which is huge. What's the broader implication for the industry in your view? 388 00:21:12,160 --> 00:21:15,880 I think the broader implication is that developers no longer have to compromise 389 00:21:15,880 --> 00:21:17,040 between speed-to-market 390 00:21:17,040 --> 00:21:21,680 Operational cost and deep security they can finally just rely on vetted 391 00:21:21,680 --> 00:21:24,400 infrastructure as we conclude this analysis 392 00:21:24,400 --> 00:21:28,200 I want to leave you with a provocative thought regarding that AI frontier 393 00:21:28,200 --> 00:21:32,860 We established that autonomous scripts AI models and micro services will soon 394 00:21:32,860 --> 00:21:35,600 completely dominate network traffic 395 00:21:35,600 --> 00:21:39,800 No doubt about it. So when a network is populated almost entirely by non-human 396 00:21:39,800 --> 00:21:42,340 agents performing automated tasks 397 00:21:42,340 --> 00:21:46,440 How will our foundational definition of digital identity have to evolve? 398 00:21:46,440 --> 00:21:50,910 Will the security frameworks of tomorrow have to assess not just the cryptographic 399 00:21:50,910 --> 00:21:52,000 signature of an AI? 400 00:21:52,000 --> 00:21:54,600 But the behavioral intent of its actions 401 00:21:54,600 --> 00:21:58,110 Honestly, that is the defining security question for the next decade of 402 00:21:58,110 --> 00:22:02,420 decentralized computing as you navigate that rapidly changing landscape 403 00:22:02,480 --> 00:22:06,320 You really don't have to be tethered to the expensive proprietary giants of the 404 00:22:06,320 --> 00:22:09,600 past and that brings us back to our sponsor safe server 405 00:22:09,600 --> 00:22:13,960 by transitioning to a robust open-source solution like log tow 406 00:22:13,960 --> 00:22:18,550 Organizations businesses and associations gain incredible cost savings while 407 00:22:18,550 --> 00:22:22,220 permanently escaping vendor lock-in, which is the dream, right? 408 00:22:22,220 --> 00:22:24,520 Whether your priority is strict data protection 409 00:22:24,520 --> 00:22:29,530 establishing immutable audit trails or simply retaining absolute sovereignty over 410 00:22:29,530 --> 00:22:30,280 your user data 411 00:22:30,680 --> 00:22:34,020 Safe server is equipped to facilitate that entire transition 412 00:22:34,020 --> 00:22:38,770 You can commission them for specialized consulting to determine if log tow or you 413 00:22:38,770 --> 00:22:39,160 know 414 00:22:39,160 --> 00:22:43,290 Another open-source alternative is the exact right architectural fit for your 415 00:22:43,290 --> 00:22:44,860 organization's specific needs 416 00:22:44,860 --> 00:22:48,990 From that very first strategy conversation the full-scale operation on highly 417 00:22:48,990 --> 00:22:50,720 secure servers right within the EU 418 00:22:50,720 --> 00:22:54,980 They manage the complexity so you can just focus on your core product visit 419 00:22:55,360 --> 00:23:01,130 www.safeserver.de to explore your options the next time you seamlessly log into an 420 00:23:01,130 --> 00:23:03,600 application take a moment to appreciate the immense invisible 421 00:23:03,600 --> 00:23:07,000 Cryptographic machinery verifying your identity and be really glad you didn't have 422 00:23:07,000 --> 00:23:09,960 to build it yourself. Thanks for joining us on this deep dive