InsightsPodcast

Episode 02

Should we knock down and rebuild the Internet Backbone again?

With Deb Radcliff, cybercrime journalist

16 September 2026 · 27 min 5 sec

Listen.

It plays here, and it keeps playing while you read the rest of the site. No account, no advertising.

The Security Question, cover art

A cyber security podcast that goes after the questions the news cycle skips, in interviews with the people who carry the same problems. Hosted by Deepayan Chanda and Neeti Rodrigues.

What was said.

Deep's own summary of the episode.

The internet was built on trust, not security — and in this episode of The Security Question, hosts Deep and Neeti trace how those early design flaws still fuel modern disasters like Mirai, Heartbleed, and billion-dollar email scams.

Joining them is veteran cybercrime journalist Deb Radcliff, who covered the landmark Kevin Mitnick case and authored the cyber thriller trilogy Breaking Backbone. She weighs in on hackers, law enforcement, personal freedom — and whether it's time to rebuild the internet from the ground up. Tune in now!

The whole conversation, in writing.

It follows the audio as it plays. Pick any time to hear that moment.

Automatically transcribed. Not yet checked against the audio.

  1. I would say the internet was never built from a security point of view, which you two both pointed out in your very well-researched introduction.

  2. The issue is it's a trusting protocol. And I remember being at Cisco Systems, I don't know, I think it was around 1999.

  3. And I was in a journalist pool listening to some of the Cisco leaders tout their latest whatever. I don't even remember what the technology was, obviously.

  4. Back then it was mostly routers and I raised my hand and I said, excuse me people, why can't we change IP?

  5. IP is inherently insecure. They laughed me out of the room, which wasn't the first or the last time I've been asked out of the room for asking very basic questions like that as a person who comes from a non-technology background.

  6. Like I said, I could... Welcome, listeners, to our today's episode of the security question. To begin with today's discussion, I would like to touch upon where it all actually started, the early design of ARPANET, which is fundamentally the precursor

  7. to the internet we see today. ARPANET was predominantly focused on enabling resilient and reliable communication within a set of trusted academic and government institutions.

  8. Let's stress upon the word trusted. This is what a mutual trust based on social or contractual or reputational trust.

  9. Security, those days were never a consideration for this trust. You know, as a network operated under the assumption that all users and systems were trustworthy, which is pretty naive in this day and age, but yeah, that was back in the day.

  10. And partly because of which there were few critical features like user authentication, access control and encryptions were not included in the foundational protocols.

  11. Now that core network protocols like TCP, IP, DNS, BGP were developed for functionality and connectivity but not for like verifying identities or securing data in transit.

  12. So this architecture is predominantly vulnerable to threats when exposed to untrusted or malicious actors as we are already witnessing these against today's internet at scale.

  13. The lack of encryption, for instance, in early protocol layers meant that data could be easily intercepted. While these were addressed as internet matured, and we have a lot more encryption protocols these days,

  14. but these are bolted on top of it. For example, HTTPS and VPNs now offer encryption, but remember that they are add-ons to the originally less secure protocol rather than integral components of it.

  15. This has contributed to a cybersecurity environment dominated by reactive measures such as firewalls, endpoint protection and inclusion detection.

  16. So these are designed as plug-in gaps, like to plug-in the gaps that we have rather than prevent them or by default security or should I say secure by default.

  17. So have you seen something like this, Neeti? Absolutely, the fact is in the, the internet was created in the early 1990s.

  18. At that point in time, it was a niche technology primarily used by academics, researchers, and government agencies. The other issue is this rise of AI automation and quantum computing.

  19. This is giving you a hint of what's coming in the next phase of the internet, one driven by intelligent networks, decentralized control. What started as a tool for a select few, if you remember, when I started about 1990s, is now

  20. an indispensable ever-evolving force driving the modern world. That said, beneath this rapid revolution, like you rightly said in the beginning, lies a whole architecture built on fragile foundations.

  21. Many of those core protocols and design decisions from the early days was never intended to support scale complexity or security, as we see in today's digital ecosystem.

  22. As a result, most of the modern day innovation is sitting atop outdated infrastructure. Patched together with temporary solutions. I think I've said enough. At this point, I'm

  23. honored to introduce a trail blazer in the world of cybercrime journalism. With a career that began in the 1990s, she was one of the first to cover what became a landmark case

  24. in cyber history. The most talked about breach involving early security researcher Kevin Mitnick. Since then, she has remained deeply embedded in the cybersecurity space, working alongside some of the brightest minds in the field.

  25. Deb, it's a true privilege to welcome you. Here you go, Deb Radcliff, who has vast and diverse experience and continues to shape how we understand and report on cyber threats today.

  26. She's put together some of her experience in the three parts, SkyFi, cyber novel called breaking backbones, which also aligns with today's security question. Should we knock

  27. down and rebuild the internet backbone again? Deb, I would like you to introduce yourself to our listeners before we start exploring our topic.

  28. Thank you. Nice to be on the show with back both of you. And yeah, I'm Deb Radcliff and I had no idea what I was getting myself into back in 1995 when I was working on

  29. a best-selling book about Kevin Mitnick who was on the run from the FBI. He was the harbinger of all that was to come. He was really good at social engineering and dissecting things

  30. and finding ways in that you're not supposed to. And his friends found me really fast and they started social engineering me and I didn't know anything about computers really except how to use them and I dove in from a general assignment newspaper reporting background

  31. at the time working for New York Times own newspaper and ended up dropping all of that for a full-time career. I like to say I am or was the first security journalist to make cybercrime

  32. a beat. We have Brian Krebs now he's great but back then I sort of was the Brian Krebs and I had to rely on hackers to understand what was going on and these are the same guys who are

  33. trying to social engineer me but they got that I was really trying to get the tech and so they worked with me really closely. But my assumption back then was this was going to be a train wreck everyone was going to be using the internet for something eventually and everyone was going to be

  34. like sheep to the slaughter regardless of whether it was going to be big businesses, small businesses, or individual users like grandma and grandpa. So I made it my mission from that

  35. point forward to try to sort of represent all of these audiences as a cybercrime journalist. Now I'm considered more of a cybersecurity journalist, but I still get to cover the juicy

  36. story every once in a while. That's excellent and impressive background there. Thanks for sharing that. You know, I'm very keen to know what your experience as a cybercrime journalist has been so far.

  37. I know we have been talking about this for last two, three years or so between yourself and myself. But, you know, from the early stages of the Internet, which you had the privilege of really witnessing from a side seat of the Internet security issues,

  38. And up until today, the modern day security issues that we see, do you think it's justified to say that internet is broken from a security point of view?

  39. I would say the internet was never built from a security point of view, which you two both pointed out in your very well-researched introduction.

  40. The issue is it's a trusting protocol. And I remember being at Cisco Systems, I don't know, I don't know, I think it was around 1999 and I was in a journalist pool listening

  41. to some of the Cisco leaders tout their latest whatever. I don't even remember what the technology was obviously back then it was mostly routers and I raised my hand and I said, excuse me people, why can't we change IP?

  42. IP is inherently insecure. They laughed me out of the room, which wasn't the first or the last time I've been asked out of room for asking very basic questions like that as a person who comes from a non-technology background.

  43. Like I said, I could see the sheep to the slaughter. I could see that the internet was built for academic situations where there was trust between both sides. I also could see that even a dog could get grandma to give money on the internet even back in 1996, 97, 98.

  44. So I had a lot of concern for the average Joe and corporations because IP was very trusting and no one could use encryption and PKI has never really done a good job and it's never been thorough enough

  45. and it's never really met its quote unquote potential. And I could go on with all the layers we've tried to put on saying the internet is broken might be a bit of a stretch, saying that internet wasn't built for what we're doing on it

  46. and we continue to innovate on top of it. I think that that has been a shortcoming in everybody's tech development. Every application we make is now where the vulnerabilities are, but the vulnerabilities are the ones that are touching the internet.

  47. So where do we go from here? That's been my question I've been bringing up at several times when I do speaking engagements and stuff, because are we gonna just keep piling everything on a technology that was built in the 80s for academics

  48. and hope that we don't absolutely totally break it someday or is it innovation? Yeah, I like the word that you pick up, right?

  49. It was meant for academic purpose, which became a mainstream requirement later on because everyone saw a benefit in connecting information systems and that's how probably it grew

  50. and all of a sudden it went out of control probably, yeah. Yeah, exactly what you're saying. And so we see more and more uses and yay that I mean, as a journalist, I can find information so much faster.

  51. I just have to go to libraries and find books and that had like five year old information on a company startup, for example, like, Oh, wait, I can't get this year's revenues.

  52. No, you can get five year ago revenues. And so things are a lot better for me as a journalist in terms of my capability to get my hands on information that I need.

  53. And that's what the internet was built for in the first place. And that's what it's good at. But the problem is, is everything else we've been talking about, the layers and layers of technology that we're putting on top of this now, and where do we go has been my question for a long time.

  54. Do we just consider IP is always going to be the foundational technology, or are we going to look at something else? And I don't see how this is going to stop right in future either.

  55. I mean, if we will keep seeing new tools, technology solutions continue to come on top of this questionable backbone from a security point of view actually we've got examples I mean I have some of them which were attacks because of early

  56. design flaws that still exists today the Mirai botnet is a classic example remember that one tens of thousands of insecure internet connected cameras this happened a way back you know 2016 which sent a huge wave of fake DNS requests to

  57. a company called DIN, at its peak, the DDoS attack was hitting 1.2 terabytes per second, taking down major websites like Twitter, Netflix, Reddit, and PayPal for hours.

  58. And this was simply because it was using a UDP protocol. So all weaknesses from the early days are still existing. I mean, Deep, do you have any examples you'd like to share?

  59. Oh yes, there are plenty of such examples, but at least few to name and notably just one attack because you mentioned about bolt-on security on top of ARPANET flaws, right?

  60. So this was back in 2014, attack called hardbleed attack, which happened because a simple coding mistake that slipped into open SSL, the very add-on software that basically creates a secure tunnel for encrypted communication.

  61. I think we all would have noticed this as a small tiny padlock on the browser address part, on any of the HTTPS websites that we visit.

  62. Now this enabled the attackers to actually copy random chunks of server memory leaking passwords, which actually caused passwords to leak through and then also the encryption keys

  63. for almost about 17% of the entire internet. Right now, this was significant because this included

  64. major websites like Yahoo and many of the US government pages and tons and tons of other websites across the globe. So this deeper problem is again, you can see in the history of the

  65. original internet, ARPANET itself, which never bid encryption into TCP/IP. So everyone depends on extra libraries for privacy like, you know, the open SSL one. Again, when that bolt-on failed,

  66. the network itself had no safety net to provide against, you know, there was no safety net to provide against this attack. Now, the same connection, so they had, you know, thought process of connect for secure later.

  67. This mindset of ARPANET shows up in many places, like in 2020, Amazon had to face a record of 2.3 terabyte per second DDoS attack that basically misused a connectionless LDAP protocol,

  68. a chatty UDP service that trusts any sender addresses is a valid address. Now, we have also seen since 2023 rise in business email compromise which we call BEC compromise these days which

  69. costs company almost over US 2 billion dollar per year because of simple mail transfer protocol which is SMTP right and this email protocol was inherited

  70. from ARPANET days which still lets anyone fake a sender you know unless optional checks like SPF, DKM, or DMARC are implemented. So these three cases that I highlight

  71. has one lesson to learn from. When security is optional, attacks are still possible in this day and age, simply because the core internet protocols were developed without security in mind.

  72. We can go on about this, Meethi, isn't it? True. And you know, as we're speaking about attacks and attackers, there's a very important question that I'd like to ask Deb. Deb, you've

  73. spent a lot of time understanding hacker communities and cyber law enforcement. So you've seen both sides of the coin. And you already gave me a glimpse about the average Joe, you just

  74. mentioned it. How do you see this ongoing push and pull between security measures and personal freedoms evolving especially since you have a very strong opinion about what an average user

  75. is entitled to. Your thoughts Deb? That's a interesting question because if you would have asked me this a year ago I wouldn't have had a good answer for you but because of what's

  76. happening today how the internet is being used against people on many levels not just in social media, but I hate to get political and I'll try to keep it minimal, but we've got governments,

  77. I won't name them, that are using social media now and our posts and things we do to start to persecute us, to try to get us to reform. We've got China who has been choking the internet

  78. for a long time, way before today's political climate, and I wrote in my cyber thriller book, this is historic, was the president, I think it was around 2017 said China needs to own all

  79. discourse on the internet and control it. And so for a while it was very freeing and it still is, especially for people who are really good at tech and they know where to get stuff on the dark web

  80. and, you know, not be monitored. They know how to use safe browsing tactics and other things. But for the average Joe, everything they're doing is pretty much being monitored. I mean, when I go buy a pair of slippers, the next time I'm using Facebook, I've got 60 slipper ads.

  81. So obviously, they're tracking my browser use. And we knew that that was the first thing I think because of the advertisers moving in and you know so it was all about advertisers

  82. but there's so much now and there's so it's not you it's not as private as anyone wants to be even the experts and we use private encryption we use encrypted email when we can okay but is any

  83. Any of this easy for the average Joe? No, it hasn't gotten easy for the average Joe. It hasn't gotten embedded in any of the applications that you're using.

  84. You have to understand this and add this to your technology stack. And you're not an IT person. Corporations do better at it, but once it gets out their door and it's going to their customers, how safe is that?

  85. So the privacy aspect, the use of technology, However we decide to use it is being monitored. And I'm also very concerned around artificial intelligence and the danger to people who may be in someone's crosshairs

  86. and they can use AI to spread misinformation about that person. They can make fake porn. They can do all kinds of stuff to try to destroy someone just using AI.

  87. So all of that is running over the internet. Again, internet protocol. And I can say that I think that the push and pull is gonna get a lot worse before and if it gets better.

  88. Yeah, so I think you brought a few very important points that because this has become so much uncontrollable in last so many years ever since the internet started and then we continue to create more and more

  89. newer applications and newer methods of doing business and communications. Somewhere this has become very chaotic and uncontrollable.

  90. So basically what I understand so far is that the initial decision to prioritize connectivity over security has forced the modern internet to operate on a very fragile and insecure foundation, at least in a large segment of the internet,

  91. if not everything, but to a majority of that internet is very, very much insecure in terms of foundational elements, right?

  92. The consequences are really significant as you rightly mentioned that we are moving towards AI and the other quantum computing stuff.

  93. So organizations actually spend billions annually to mitigate these threats that emerge from these legacy design choices and the internet users remain exposed, like you rightly said,

  94. to risks like maybe malware, phishing, ransomware, or any kind of scams that's happening every single day. The complexity of retrofitting security into insecure by default systems, I really call it insecure by default systems,

  95. results in issues and actions like misconfigurations or human error. Further increases the problem, right? Now, this historic, with this historical context is what brings me to the next question for you, Dev.

  96. Do you really think that we should break the internet and rebuild it again from ground up? What does the cybercrime journalist inside you think about it every day, or maybe you have been thinking about it ever since?

  97. Well, I've been hoping, but I haven't really found a good answer to what's being done but I've been hoping that we would see innovation around a new

  98. communications and transportation infrastructure that would be different than TCPIP and that would run in parallel until we've worked out all the kinks and all the problems and all the things

  99. that could go wrong and then once we're sure that it's a much more secure, much more stable, and much more expandable and resilient infrastructure, we move over to that.

  100. But of course, no one's gonna wanna change their applications. Everything's based on IP, every app that runs on the internet, every app that talks to the internet, every API, every backend database, they're all programmed to reach out to an environment over internet protocol,

  101. which means completely redoing all the apps that we're used to today, completely redoing everything and I don't know what it would take for that. Now we're looking at quantum computing and you guys brought that up in the beginning and that's more of an organic

  102. from what I can tell form of computing and but it's about qubits and it's about it can only go you know 15 miles before it runs out of steam and we need something global. We know that we're

  103. sending man-missaged Mars someday. How are they going to communicate with those missions? Right now, a lot of space stuff is still being done with applications that are built over IP

  104. for communications and other things. We've got old code like C and C++ that are the only ones that work well for these type of applications today. So how are we going to take all of that

  105. and change it. And that's why I got laughed out of the room with Cisco. Even back in 1999, when I asked the question, they said there's way too much invested in this. It's not going to change.

  106. Well, we've upgraded to IPv6, or some of us have. So IP has changed a little, okay, but will it support all types of future applications that we want to do? And will it support it safely? And if

  107. not, then why are we not looking at a parallel transport technology? And if I'm hoping maybe someone in this audience will tell us, hey, by the way, there is this out there, because I've done a lot of looking because I've done a lot of speaking on this. And the closest I can come to is Quantum.

  108. And Quantum is resource heavy. Quantum doesn't go very far. Quantum may or may not be a better solution for, at least for internet transport. Yes, we can use it for better encryption,

  109. but I'm talking about IP. So I don't think we should break it and start over. I think we should build something in parallel. And then when we've got it all worked out, and we've got buy-in from

  110. all the application vendors that are laying on top of this, and they've got the protocol straight, and they've got everything they need and best practices, and we're ready to go, then we should be rolling out to a new infrastructure.

  111. You guys, that's not going to happen in our lifetime. But that's my thought, and that's been my thought all along. If it's broken, why aren't we fixing it? Yeah, I think, you know, you brought up very important points, Dev, and nicely put, actually.

  112. So, well, with that, was a very nice and engaging discussion with you today, Dev and Niti. As the industry rushes towards the era of quantum computing and AI,

  113. there is no doubt that we will need a new and solid security foundation or maybe a parallel one like you rightly mentioned there.

  114. And the era of unsecured internet has to come to an end. And I will pause our discussion today with the hope that someday the industry will take a bold move

  115. a bold move and either rebuild a secure internet from its core foundation or create a parallel one which is more secure and also hope to get another round of exciting discussion with you

  116. Dev in future when your cyber war trilogy becomes a Netflix web series and really hoping that it becomes a Netflix web series one day in near future. So we will have another opportunity

  117. to talk to you. Until then, thanks to both of you and have a great day ahead. Thank you, Deb. And looking forward to the Netflix series.

The other episodes.

  1. Episode 01,

    47 min 35 sec

See every episode

Start with a conversation.

If an episode touches a decision in front of you, a short conversation is the quickest way to test it against your own situation.

Book a conversation