


Election 2026 cyber threats and misinformation; voters’ backlash against data-center expansion tied to AI; industrial refrigeration hacks; OpenAI security warnings; NVIDIA China compute restrictions; enterprise AI/agent governance and “agent sprawl”; plus assorted security news (Binance hack, Barracuda flaws, SOC staffing trends, Canadian university hacking indictments).
Heather Jalen, CISO at Box. Background: leads security at Box; focuses on enterprise AI governance, runtime guardrails, identity lifecycle for agentic systems, and observability/assurance for AI agents.
AI agents shift risk from model security to runtime governance; permissions must be fine-grained and enforced in runtime; high-risk actions require human-in-the-loop; agent sprawl mirrors SaaS sprawl and requires visibility and risk-based prioritization; agent identities need lifecycle management and downscoped credentials.
Westchester County ballot scanner malfunctions fueling X conspiracy posts; refrigeration attacks changing settings/alarms (including French blood freezer settings); Reuters claims OpenAI employees warned about repository access and security reviews; Binance used compromised engineers’ machines and an internal AI tool to reach wallet signing; Box discusses agent buckets (low/mid/high risk) and “pressure testing” lessons.
Written by AI. May contain mistakes.
Discussing the increasing stakes and challenges in cybersecurity for fast-scaling companies.
“When you run security for a company that's scaling fast, the stakes just keep climbing.”
Analyzing cybersecurity threats and issues during Election Day 2026.
“Gartner predicts an engineering-heavy future for security operations.”
Exploring voter sentiments regarding AI and its impact on data centers.
“Election Day 2026 brought a familiar mix of cybersecurity threats, technical glitches, and misinformation.”
Discussing recent cyber attacks affecting refrigeration systems across Europe.
“but it turns out the people living next door to those data centers get a vote, too.”
Overview of security challenges faced by OpenAI as revealed by former employees.
“changing freezer settings from minus 30 degrees Celsius to 110 degrees.”
Insights into NVIDIA's limitations on data center customers regarding China.
“The larger challenge is protecting increasingly valuable AI technology without slowing the research that makes it valuable.”
Details on Qualcomm's partnership with SpaceX for a new 5G network.
“She joins us each week with the latest space cyber news.”
Report on the indictment of two Canadians for hacking U.S. universities.
“Be sure to check out the T-minus Space Cyber podcast wherever you get your favorite shows.”
Learn about the allegations against cyber criminals targeting universities.
“Prosecutors say Ryan James Roach and Matthew James Stubbings targeted institutions in four states between 2016 and 2019, deploying malware, stealing credentials, and exploiting university supercomputers.”
Learn about the allegations against cyber criminals targeting universities.
“Mimecast still protects your email, and that is not changing.”
Explore how AI is transforming business operations and security risks.
“2026 has been a very transformational year for AI, just AI in general.”
Discuss the importance of managing AI agents and the risks involved.
“And as these agents are becoming more powerful, you really have to think through what are the guardrails I need?”
Understand how to manage identities and access for AI agents.
“I think if you think about SaaS sprawl and agent sprawl, I mean, we're kind of reliving the SaaS sprawl that we did back in the early 2000s, just at a much faster and I think much larger scale.”
Learn about the challenges of proving AI governance in organizations.
“So in those instances, what we're trying to do is, you know, downscope the access for that agent.”
Explore the considerations for building or buying AI solutions.
“a small bit of comfort that everyone's trying to figure that out at the same time, right?”
Exploration of the trade-offs between developing in-house AI tools versus outsourcing to third parties.
“The trade-off you have to weigh is, do I want my team building agents and running these tools and maintaining them?”
An introduction to SpecterOps' Bloodhound Enterprise for securing AI-driven identities.
“And now a word from our sponsor, SpecterOps.”
Promotional segment for Meter, emphasizing their network solutions.
“Yes, you can have an enterprise network that's secure and reliable and high performance.”
Ads for Google Chrome's new features and Carhartt's holiday gift guide.
“This episode is brought to you by Google Chrome.”
Discussion on the use of AI in managing fantasy soccer teams and its effectiveness.
“And finally, Wired's Will Knight decided to see whether artificial intelligence could outperform him at fantasy soccer.”

Get the gist, explore an idea or ask about something said.
On CyberWire Daily, what are the key points of this episode?
Automatic transcript. May contain errors.
0:01You're listening to the Cyber Wire Network, powered by N2K.
0:12When you run security for a company that's scaling fast, the stakes just keep climbing. More compliance frameworks, more vendors, more risk, and a board that wants to see it all in one place. And all the information you need, it's scattered everywhere. Meanwhile, you're making this quarter's decisions based on last quarter's data. Vanta offers a better way. Vanta's Agentic Trust platform connects your compliance program with over 400 integrations and continuous monitoring. It automates evidence collection, helps identify the risks that deserve your attention, and organizes audits around your auditor's requirements.
0:51And the Vanta agent works around the clock with the full context of your program. Vanta reports a 526 % return on investment over three years, paying for itself in just three months. It's trusted by more than 16 ,000 companies, including Snowflake, Atlassian, and Ramp. Spend less time chasing compliance data and more time managing risk. Learn more at vanta.com slash cyber. That's V-A-N-T-A dot com slash cyber.
1:30This episode is brought to you by Google Chrome. You think you know a browser, but Gemini and Chrome, that's new. It can help you with practically anything on the web, like restoring a vintage motorcycle from a 50-page restoration block, or finally break down that long article you've had open for weeks. Gemini and Chrome is here for it. Ready to make anything online make sense? There's no place like Chrome. Check responses set up required, compatibility and availability varies 18+.
2:23Thank you. security flaws. Gartner predicts an engineering-heavy future for security operations. Canadians face charges for hacking U.S. universities. Our guest is Heather Jalen, CISO at Box, discussing enterprise AI in 2026, what security leaders learned, and what comes next. And AI takes the pitch and misses the point.
3:00It's Wednesday, November 4th, 2026. I'm Dave Bittner, and this is your CyberWire Intel Briefing.
3:21Thanks for joining us here today. It is great, as always, to have you with us. Election Day 2026 brought a familiar mix of cybersecurity threats, technical glitches, and misinformation. But the biggest technology story emerging from the midterms may be voters' growing resistance to artificial intelligence. According to NextGov, election officials reported unsuccessful attempts to penetrate online services, including efforts to guess login credentials. One state election office experienced five spikes in traffic to its ballot dropbox locator, which subsided after officials implemented a CAPTCHA.
4:01As of Tuesday morning, election security analysts had received no reports of foreign influence operations. Meanwhile, ballot scanner malfunctions in New York's Westchester County quickly became fodder for election conspiracy theories. Officials initially attributed rejected ballots to problems with their thickness. Voters were instructed to deposit affected ballots in secure emergency bags for later counting. Despite no evidence of deliberate interference, misleading claims spread rapidly on X, reaching as many as 100 posts per minute. Beyond election security, voters delivered a significant political setback to the technology industry's ambitious AI expansion.
4:46Axios reports that roughly two dozen ballot measures restricting data center development appeared headed for approval. In Texas, Democratic gubernatorial winner Gina Hinojosa campaigned on maintaining a construction moratorium until lawmakers address environmental and infrastructure concerns. With Democrats also winning key races in Ohio, Georgia, and Arizona, analysts estimate nearly two-thirds of planned U.S. data center capacity for 2028 could face delays. The emerging message for voters seems to be they're increasingly concerned about AI's impact on electricity prices, water supplies, and their communities.
5:29Silicon Valley may have enormous ambitions, but it turns out the people living next door to those data centers get a vote, too. Hackers are targeting refrigeration systems across Europe, remotely changing temperature settings and forcing supermarkets to throw away spoiled food. Attacks reported in November affected stores in Sweden, Belgium, Luxembourg, Ireland, and Germany, along with food suppliers in the UK. In some cases, attackers switched refrigeration equipment into defrost mode or disabled temperature alarms. Retailers have responded by disconnecting vulnerable equipment from the Internet.
6:10The attacks have also raised concerns about medical supplies. In France, hackers reportedly compromised refrigeration equipment supplied to the French blood establishment, changing freezer settings from minus 30 degrees Celsius to 110 degrees. An alert technician caught the problem before it disrupted operations, according to French health officials. There's no confirmed loss of blood plasma or harm to patients. Authorities haven't identified the perpetrators or established whether the incidents are connected. French cybersecurity officials are urging organizations to secure remote access to refrigeration controllers, update firmware, and eliminate default passwords.
6:53The incidents demonstrate how compromising ordinary industrial equipment can have very real physical consequences. Current and former OpenAI employees say they warned company leadership for years about security weaknesses that could expose sensitive AI research to competitors and foreign governments. According to Reuters, concerns included broad employee access to internal code repositories, sensitive research documentation, and inadequate security reviews of internal tools. Some employees believed leadership prioritized development speed over security. Those concerns gained urgency following several incidents, including the arrest of a former OpenAI engineer accused of planting a listening device in an executive conference room and researchers using a rival AI model to access employee accounts and internal systems.
7:51OpenAI disputes suggestions that it neglected security, pointing to restricted repositories, increased investment, and regular reviews. The company reportedly reassigned a quarter of its production engineers to security improvements in August. The larger challenge is protecting increasingly valuable AI technology without slowing the research that makes it valuable. NVIDIA is reportedly imposing new restrictions on data center customers that provide computing services to Chinese companies. According to the Financial Times, NVIDIA has told operators they must limit Chinese customers to no more than 49 % of their overall business, with the remaining 51 % coming from non-Chinese companies.
8:38The policy would require data center operators to track how much business they conduct with Chinese firms, potentially complicating operations for providers serving customers in both China and elsewhere. The Financial Times says NVIDIA's sales representatives and regional teams have communicated the restrictions directly to customers, but the company apparently hasn't issued formal written documentation. The reported move introduces a new limitation on how NVIDIA-powered computing infrastructure can be used, adding another consideration for data center operators navigating business relationships with China.
9:19Maria Vermazis is host of the T-Minus Space Cyber podcast. She joins us each week with the latest space cyber news. Thank you, Dave. Wireless technology and semiconductor company Qualcomm signed a multi-year deal with SpaceX to supply equipment and software for Starlink Mobile's planned nationwide U.S. 5G network. Qualcomm described the deal as its largest 5G infrastructure deal to date, and the agreement will cover tens of thousands of cell tower sites, with deployment expected to begin in 2027. And this is just the latest move as SpaceX seeks to challenge established wireless carriers in the United States like AT &T, Verizon, and T-Mobile.
10:07The SpaceX wireless network will combine terrestrial 5G with satellite connectivity. Qualcomm's technology will support radio communications, baseband processing, and advanced beamforming, which directs wireless signals from satellites towards users on the ground in order to improve coverage and make efficient use of RF spectrum. The initial rollout of the equipment and software deal between Qualcomm and SpaceX will focus on terrestrial 5G, but the architecture is being designed to accommodate future connectivity and network intelligence developments. For the Cyber Wire Daily, I'm Maria Varmazes from T-Minus Space Cyber Briefing.
10:45Back to you. Be sure to check out the T-minus Space Cyber podcast wherever you get your favorite shows. North Korea has launched several missiles toward the sea, hours after rejecting FBI accusations that its hackers stole more than$6 billion from cryptocurrency exchange Binance. The FBI attributes the theft to North Korea's Trader Trader Group. Pyongyang denies responsibility and says the accusations undermine prospects for renewed diplomacy with Washington. Meanwhile, Binance has revealed how the attack unfolded. Hackers spent months compromising three engineers' computers using malicious software disguised as developer tools and a fake job interview featuring a rigged coding assignment.
11:36Attackers also exploited an internal AI tool to expand their access, eventually reaching Binance's wallet signing infrastructure. The company hasn't explained precisely how the AI tool was abused. Binance says approximately$2.85 billion in stolen assets has been frozen or intercepted, and affected systems have been rebuilt. Barracuda Networks is urging customers to patch two vulnerabilities in its email security gateway appliances that could allow attackers to execute commands with root privileges. One critical flaw enables unauthorized changes to global configurations, while the second allows users with administrator permissions to inject operating system commands.
12:23Barracuda says attackers must first reach the device's administrative interfaces. Chaining the vulnerabilities could allow them to bypass authorization controls and gain root access. The company has released patches and recommends updating firmware, applying security fixes, and restricting administrative access to trusted internal networks. By 2028, security operations centers could employ 50 % more engineers than analysts, according to a new Gartner report. The shift reflects AI's growing ability to handle routine security alerts, freeing human analysts to focus on investigating cases, making decisions, and coordinating responses.
13:08Gartner predicts SOCs won't necessarily shrink. Instead, engineers will increasingly build, test, and refine AI workflows while analysts expand into threat hunting and exposure management. The transition creates challenges for entry-level analysts who traditionally learned their craft by processing alerts. Organizations will need new approaches to training and developing expertise. Gartner also recommends replacing traditional alert volume metrics with measures of case accuracy and decision quality By 2030, the firm expects SOX to become more proactive Continuously testing their own defenses rather than primarily responding to alerts Two Canadian men have been indicted for allegedly hacking U.S.
13:57universities to steal data and mine cryptocurrency Prosecutors say Ryan James Roach and Matthew James Stubbings targeted institutions in four states between 2016 and 2019, deploying malware, stealing credentials, and exploiting university supercomputers. They allegedly used servers registered under false identities and compromised systems to conceal their activities. Roach has been extradited to the United States while Stubbings awaits extradition. Both face up to 10 years in prison and fines of$250 ,000 if convicted.
14:44Coming up after the break, my conversation with Heather Jalen from Box. We're discussing enterprise AI in 2026. And AI takes the pitch but misses the point. Stay with us.
15:09Heather Ceylan: Mimecast still protects your email, and that is not changing. But they haven't stood still. They've built around the person behind the inbox, across Slack, Teams, the browser, and the AI agents acting under their login. 8 % of employees drive 80 % of the risk, and now that includes their agents. Mimecast Apex is one platform that shows you where your risk really sits and protects the work wherever it happens. The Mimecast you trust with email now covering far more. Mimecast secures humans, data, and AI. Start at Mimecast.com.
15:50This episode is brought to you by Google Chrome. You think you know a browser, but Gemini and Chrome? That's new. It can help you with practically anything on the web, like restoring a vintage motorcycle from a 50-page restoration block, or finally break down that long article you've had open for weeks. Gemini and Chrome is here for it. Ready to make anything online make sense? There's no place like Chrome. Check responses set up required, compatibility and availability varies 18+. traditional home security only alerts you after a break-in and that's too late simply safe is changing that stop this is simply safe police are on the way we don't just alert we help stop crime before it starts plan starting around a dollar a day right now get 70 off your new system with professional monitoring at simply safe.com slash spotify simply safe ahead of crime in real time outdoor deterrence plans start at 49.99 a month visit simply safe.com slash licenses for alarm license information.
16:45When you're building a company, you learn that trust is what closes deals. You may have the best product, but buyers won't sign without proof that your product is secure, especially if it has AI. Vanta gets you compliant fast with SOC 2, ISO 27001 and more, and keeps you secure with continuous monitoring. Hyperbound, a YC-backed startup, used Vanta to unlock the Fortune 100 in under a year. Snowfire got audit ready in a month and closed a multi-million dollar deal. Trust matters. Get started at Vanta.com.
17:20Heather Jalen is Chief Information Security Officer at Box. And in today's sponsored Industry Voices conversation, we discuss enterprise AI in 2026, what security leaders learned, and what comes next. 2026 has been a very transformational year for AI, just AI in general. And then also from a security perspective, I think we're starting to see the transition from people using AI as assistants to people using AI to actually transform business processes and how they operate day to day. So I think that agents are becoming more and more ingrained in our workflows today, which is great because it's making us more productive, but it also introduces a lot more both security and resiliency risks that we have to prepare for and think about.
18:11Heather Ceylan: There's been a lot of attention paid to securing the AI models themselves. I know you've emphasized that governance really needs to begin before the model with the information it's accessing. Can you walk us through what that means in practice? The way I like to think about it is agents, they're going to be at their best when they're operating off enterprise content. And they have the context from your business. If you look at, you know, AI models in general, they're trained on very general public data. Once you give them enterprise context, then they're much more powerful for transforming how you work.
18:46And so in order for them to get that enterprise context, what's really important is making sure you have those controls around your content and you have those fine-grained permissions so agents can only access what you intend for them to access. we've always kind of struggled to get those fine-grained permissions controls with humans and now with agents you know humans can pause think be held accountable agents can transverse you know terabytes of data in in seconds so that's a new reality we have to control for as we're thinking about opening our enterprise content up for agents and a lot of that is moving security into the runtime and putting guardrails in the runtime and not just solely relying on access controls.
19:36Heather Ceylan: Yeah, it seems like a potential for unintended consequences when we open the AIs up to discover, analyze, and even act upon decades of corporate documents and files and all this unstructured data. It's a very scary thing to think about, but it's also, I don't want to be all about fear, it's also super powerful, right? Like you want those agents to have that context. But I think you have to, there are very real consequences. And as these agents are becoming more powerful, you really have to think through what are the guardrails I need? And what's the worst case scenario of what could go wrong? And how do I prevent that?
20:19When I think about, you know, how we think about agents at bots, as we're deploying them across our enterprise, I think about them in three buckets. There's those bucket of agents that are, you know, very low risk, low consequence. They're doing, you know, metadata extraction, summarization. If they get something wrong, it's not going to break anything major downstream. So those you can kind of automate freely. Then you have kind of a middle bucket where, you know, maybe it's integrated into a workflow and it could give a feed bad data into a workflow that could have downstream consequences. that maybe you want to automate with some guardrails and putting those guardrails in place in the agent runtime itself.
21:01And then there's this third bucket where it's like, we're always going to have a human in the loop because this is such a high risk action, like publishing something to the web or publishing something, making it publicly available, sending something out the door, you know, for our security operations center, right? Like making a final call and taking an action on an incident. Those are things where we will always have a human in the loop because those actions are irreversible.
21:30Heather Ceylan: I know one of the things you've identified going forward as a concern is this notion of agent sprawl. What do you mean by that? And how is that different from, say, shadow IT? I don't think it's very different at all, actually. I think if you think about SaaS sprawl and agent sprawl, I mean, we're kind of reliving the SaaS sprawl that we did back in the early 2000s, just at a much faster and I think much larger scale. So it makes it much more difficult to control. And I think what a lot of organizations are dealing with right now is, you know, people are being encouraged to experiment with agents and that's great.
22:11and we have a lot of agents that are you know growing out of experiments and you know we're testing a bunch of different things and so how do you you know productionize those agents right and how do you figure out which agents are going to be a part of your organization and which ones you're going to spin down and I think organizations are still really trying to get a handle on that and a lot of organizations that I talk to are still trying to figure out what is even running in my environment. You know, that's the first rule. That's nothing new in security. You can't govern what you can't see.
22:49So making sure you're at least getting, you start by getting the visibility into the agents that are running in your environment. And then from there, you just have to prioritize based on risk because you're not going to be able to tackle everything at once. You have to look at what are my highest risk agents based on the data they're touching, the actions that they're performing, the access they have, and then work on controlling those first.
23:13Heather Ceylan: What about identity and access management? You know, I'm thinking of an agent that outlives its original purpose or, you know, somebody moves on and nobody remembers who created that agent that's running around inside the organization. How should organizations approach that life cycle management with these machine identities? That's a very real risk. And it's a very real problem that we're seeing across the industry already. And we're still pretty new in terms of the adoption of agents, you know, just in general. And we're already seeing challenges governing the identity. So, I mean, they have to be managed like human identities, but in different ways.
23:57So the same principles need to apply, right? You need to understand each agent, each identity it has, its purpose, the challenges with some of these, you know, more broad scope agents that are operating off these broadly scoped credentials, those are much harder to control, right? And much harder to give the fine grain permission. So in those instances, what we're trying to do is, you know, downscope the access for that agent. So if the agent is being asked to review this document and summarize it and put it in this folder, scoping the access privileges so it only has that access. It can't go make that file public.
24:39It's not even like in the agent's vocabulary to call. So I think as we start to, you know, deploy more and more agents across the business, focusing on how you govern those identities and how you manage them, they're kind of like a mix between identities and applications. And you have to, you know, you have to manage both. And then you have to really make sure you're not giving overly scoped access and you're trying to downscope the access grants to match the tests that the agents get in.
25:10Heather Ceylan: Well, you mentioned governance. I know one of the things that you're focused on, again, looking forward is that there's going to be a point when organizations need to prove that their AI is governed, how are they going to do that? How does a proof play out? That's a great question. I don't know that I have a fully solid answer to predict the future, but I think there's a couple of ways that at least we're thinking about doing that now. You have to, again, start with just the visibility and making sure you have that visibility, then you get into observability. So not just, you know, logging what action an agent took, but logging all the reasoning steps along the way.
25:58And then making sure that you can go back and recall, you can adjust the decision that the agent made. So you can really control the impact if anything gets misaligned. And I think starting there is a good starting place. And then you have to translate those metrics into something, those observability metrics into something that provides some type of assurance. And you start thinking about, you know, key risk indicators that you can tie to agent activity. For example, like, do we have agents that are performing actions that are maybe not authorized, or we have agents that are actually misaligned with the task and how many instances of that do we have?
26:41And then once you start to get that data, you can start surfacing that and monitoring and starting to see trends. But I think it's really hard to define right now what, you know, agent governance looks like and what the ideal state of agent governance is, because I think that's still very much an evolving space. Yeah.
27:00Heather Ceylan: I mean, I guess it's a small bit of comfort that everyone's trying to figure that out at the same time, right? There's nobody's ahead of the curve there. Yeah. And, you know, one organization will say, We have agent governance. Probably every organization has some form of agent governance, or most organizations now, but it's going to look different everywhere you go. So I think we're still trying to align on what that single framework looks like. What about this notion of build versus buy? I think a lot of enterprises are struggling with whether to try to build their own AI capabilities or just adopt platforms from established providers.
27:41Heather Ceylan: Looking at that from a security perspective, what sort of things should drive that decision? I think it depends on your organization. If you are a very engineering-first organization, in a lot of instances, you're going to want to build your own agents that run across multiple platforms. So at Box, we're very much an engineering-first organization. We like to build things. The trade-off you have to weigh is, do I want my team building agents and running these tools and maintaining them? Or do I want to pay a third party to do that? And there's always trade-offs, right? You lose control if it's with a third party.
28:23If it's in-house, you can constantly make changes, make updates the way you want. Especially as the models change, you can adjust your agent based on the underlying model it's using. So I think for us, we're definitely taking a very build first approach. But there are platforms where we have, you know, like SaaS applications that have agents deployed where we're leveraging the, you know, the agents in the SaaS platform itself. So I think it's, it's really just a business decision. I have a lot of peers that I talked to that, you know, they're not very engineering first organizations. And so they're buying everything, right?
29:03And that probably makes the most sense for those types of organizations.
29:08Heather Ceylan: Yeah. It seems to me like, you know, a big part of our conversation today comes down to trust and trust is a competitive advantage. Do you see it being a differentiator between organizations? Absolutely. I think, especially for organizations that are trying to sell or deploy agents as a part of their software, I think trust is a massive factor. And I think the more companies can be transparent about how their agents operate, how they're controlled, customers want to understand that it's... They want to understand it. They don't want to just... It can be scary when it just seems like this black box of, you know, I don't really understand what this agent's doing or why it made the decisions it's made.
29:52So I think the more companies can be transparent and again, give that observability to their customers around what the agent's doing and why, and then giving them control and guardrails around those agents, that goes a long way for building trust.
30:08Heather Ceylan: When you look back at this year, are there any fundamental lessons that you think security leaders have learned so far? we're learning the lesson right now that the um the next gen vulnerability discovery from these frontier models is definitely accelerating and i think we're going to see more and more instances of these massive you know wide impacting vulnerabilities like log4j the php vulnerability we recently um we're all dealing with so that time is coming it's it's here now and we're not like preparing for it anymore we are actually actively in it so i think that was that was a lesson for us and then i also think the agent identity and agent pressure testing has been has been a really big lesson for us this year it's hard to like it's hard to imagine that like the whole thing that happened with mythos was like in april of this year because it feels like so long ago.
31:11Heather Ceylan: Right. And it's just, yeah, so much changes in one year. It's going to be crazy to, you know, sit here in the same seat next year and be talking about completely different concepts, probably. That's Heather Jalen, CISO at Box.
31:43Heather Ceylan: And now a word from our sponsor, SpecterOps. Today, AI is rapidly adding non-human and agentic identities to modern enterprise environments, creating new trust relationships and attack paths. Bloodhound Enterprise helps defenders map attack paths across AWS and hybrid environments as one connected graph, identify the choke points that matter most, and bring trusted attack path intelligence into approved AI workflows with Bloodhound Hunter. See how SpecterOps helps teams secure the AI-driven identity era at specterops.io.
32:34Yes, you can have an enterprise network that's secure and reliable and high performance. And no, you don't need to choose the best two out of three. With Meter, you can get the end-to-end network built from the ground up, fast to deploy, and easy to manage. That's because Meter is software-led for easy installation, maintenance, and control for everything running on your enterprise network. Hardware, firmware, and software all working together from the start seamlessly on a unified platform that's secure by design. You can't protect what you don't know exists, which is why Meter gives you comprehensive visibility into wired and wireless routing, switching, firewalls, DNS security, and VPNs.
33:19you'll really know what's running on your network down to the most granular client level. Step off the hardware box upgrade treadmill and switch to Meter for a predictable fee and free up your team to spend time on all the other things that keep your business running. Try it out for yourself and book a demo online at meter.com slash cyberwire. That's meter.com slash cyberwire.
33:54This episode is brought to you by Google Chrome. You think you know a browser, but Gemini and Chrome? That's new. It can help you with practically anything on the web, like restoring a vintage motorcycle from a 50-page restoration block, or finally break down that long article you've had open for weeks. Gemini and Chrome is here for it. Ready to make anything online make sense? There's no place like Chrome. Check responses set up required. Compatibility and availability varies 18+. Carhartt's holiday gift guides were built in collaboration with the greatest gift giver of all time, Santa Claus himself.
34:27Old St. Nick counts on Carhartt's durable gear to have his back through rain, sleet, and snow on the biggest night of the year. Because when everyone's counting on you to do your job, what else would you wear to work? Tap the banner to grab a gift for every hardworking person on your holiday shopping list or head to Carhartt.com.
34:53And finally, Wired's Will Knight decided to see whether artificial intelligence could outperform him at fantasy soccer. He assembled a team of AI agents, handed them control of his lineup, and waited for the tactical brilliance to begin. Their opening move? Benching Erling Haaland, one of the world's most prolific goal scorers. Not exactly an auspicious start. The AI eventually reconsidered, but its subsequent decisions haven't been particularly impressive. After several weeks, Knight's AI-managed team still hasn't outperformed the average fantasy squad. Meanwhile, researchers studying AI forecasting have found that leading models can sometimes outperform ordinary people at predicting events, although expert human forecasters generally remain formidable competition.
35:46Some experiments suggest AI is closing that gap. Knight suspects AI could eventually become an unbeatable fantasy manager, but he's not particularly enthusiastic about that prospect. After all, fantasy sports are supposed to reward human judgment, even when that judgment is spectacularly wrong.
36:18And that's The Cyber Wire. For links to all of today's stories, check out our daily briefing at thecyberwire.com. We'd love to know what you think of this podcast. Your feedback ensures we deliver the insights that keep you a step ahead in the rapidly changing world of cybersecurity. If you like our show, please share a rating and review in your favorite podcast app. Please also fill out the survey in the show notes or send an email to cyberwire at n2k.com. N2K's lead producer is Liz Stokes. We're mixed by Trey Hester with original music and sound design by Elliot Peltzman. Our contributing host is Maria Vermazis.
36:56Our executive producer is Jennifer Ivan. Peter Kilpie is our publisher. And I'm Dave Bittner. Thanks for listening. We'll see you back here tomorrow.
37:12Thank you.
37:42Security helps organizations control every identity with zero-trust policies that govern what they can access, what they can do, and when. Because when critical systems are at stake, compromised access can mean downtime, disrupted production, and real-world consequences. Protect what matters most with the Zage Identity Fabric platform. Learn more at Zage.com. That's X-A-G-E dot com.
38:16Black Friday, well shop, Wayfair now, month-long big deals, yeah these are steals, 80 % less money spent. You can't go wrong, check out the song, these deals go more, black Friday scores, your most is two, more than a few. Grab a big bag, two layer head mattresses, two sheets that are new, save that Wayfair, save on these chairs, find just your style, it makes you smile. Here's a quick plug, area rose, free shipping while Wayfair is how, win the season, win the win the season. Wait there, wait there, wait there, wait there, wait there. Does checking your bank statements feel like digging through a dusty attic full of forgotten subscriptions and old free trials?
38:54Experian can help take the pain out of canceling by handling it for you. Think of it as Operation Get Your Sanity Back. Over 200 subscriptions are cancelable. So get started in the Experian app and see how much you could save. Results will vary. Not all subscriptions eligible. Savings not guaranteed. Paid membership with connected payment account required. to see Experian.com for details.
Election Day brings typical cyber threats, technical glitches, and misinformation. Hackers tamper with European refrigeration systems. OpenAI faces questions over years of security warnings. Nvidia reportedly restricts business with Chinese data center customers. Binance details a six-billion-dollar heist attributed to North Korea. Barracuda patches critical security flaws. Gartner predicts an engineering-heavy future for security operations. Canadians face charges for hacking U.S. universities. Our guest is Heather Ceylan, CISO at Box, discussing Enterprise AI in 2026: what security leaders learned — and what comes next. AI takes the pitch, misses the point.
Remember to leave us a 5-star rating and review in your favorite podcast app.
Miss an episode? Sign-up for our daily intelligence roundup, Daily Briefing, and you’ll never miss a beat. And be sure to follow CyberWire Daily on LinkedIn.
CyberWire Guest
On today’s Industry Voices segment, we are joined by Heather Ceylan, CISO at Box, discussing Enterprise AI in 2026: what security leaders learned — and what comes next. If you enjoyed this conversation, be sure to check out the full conversation here.
Selected Reading
Ballot scanner malfunctions, flurry of federal legal actions test election officials on final day of voting (CyberScoop)
Voters deliver "AI earthquake," rebuking tech and billionaires (Axios)
Election office reports failed attempts to hack its online services (Nextgov)
Hackers Hijacked French Blood Plasma Freezers, Changing Temperature Settings From −30°C to 110°C (Cyber Security News)
Hackers Alter Refrigeration Settings, Forcing European Stores to Discard Food (Hackread)
OpenAI employees warned about operational security lapses for years, sources say (Reuters)
Nvidia tells data center operators to limit business from Chinese customers with new global restrictions, report claims — allegedly restricted to 49% of the total, ensuring 51% are non-Chinese outfits (Tom's Hardware)
Qualcomm signs major 5G infrastructure deal with SpaceX for Starlink Mobile network (Reuters)
Rigged Coding Assignment, Abused AI Tool Opened Path to $6 Billion Binance Heist (SecurityWeek)
North Korea Launches Missiles Toward the Sea as Pyongyang Protests Against US Claims on Binance Hack (SecurityWeek)
Barracuda warns of critical flaws allowing root access to devices (Bleeping Computer)
Gartner Predicts SOCs Will Have 50% More Engineers Than Analysts by 2028 (Bleeping Computer)
US charges two Canadians accused of hacking US universities (Bleeping Computer)
I Put a Bunch of AI Agents in Charge of a Fantasy Soccer Team. They’re Not Better Than Humans—Yet (WIRED)
Share your feedback.
What do you think about CyberWire Daily? Please take a few minutes to share your thoughts with us by completing our brief listener survey. Thank you for helping us continue to improve our show.
Want to hear your company in the show?
N2K CyberWire helps you reach the industry’s most influential leaders and operators, while building visibility, authority, and connectivity across the cybersecurity community. Learn more at sponsor.thecyberwire.com.
The CyberWire is a production of N2K Networks, your source for strategic workforce intelligence. © N2K Networks, Inc.
Learn more about your ad choices. Visit megaphone.fm/adchoices