The CyberWire Daily Podcast 11.16.26
Ep 2678 | 11.16.26

Flightageddon grounds the nation.

Transcript

U.S. airline operations return to normal. Microsoft unpacks last week’s outage. Commerce Secretary Lutnick defends less stringent export controls on advanced AI chips. Italy fines Meta 887 million dollars for privacy violations. Telegram’s new widgets may deliver a RAT. The FBI warns of ARTEX pen testing abuse. Refrigeration controllers see opportunistic scanning. Publicly available AI models dramatically accelerate vulnerability discovery. Monday business briefing. Our guest is Patrick Coughlin, CEO and Co-Founder of Savi Security, sharing how shopping fraud has industrialized in the AI era. A refrigerator company learns it’s secretly leading the laundry revolution. 

Today is Monday November 16th 2026. I’m Dave Bittner. And this is your CyberWire Intel Briefing.

U.S. airline operations return to normal. 

U.S. airline operations are largely returning to normal following a rare four-hour nationwide ground stop issued by the FAA on Friday, November 13, which caused thousands of cancellations and stranding travelers into the weekend.

Officials confirmed the disruption was not caused by a cyberattack, attributing it instead to aging telecom infrastructure and legacy hardware. Specifically, an L3Harris contractor upgrading copper wire to fiber replaced a Cisco 600-series router manufactured between 2001 and 2006. The faulty router began flooding the network with login attempts, spreading instability across 17 FAA facilities and prompting the agency to halt flights out of an abundance of caution while 3,500 planes were airborne.

In response, the Trump administration increased its funding request for air traffic control modernization to $30 billion, noting that a previous $12.5 billion allocation is insufficient. However, Democratic lawmakers pointed out that only $1 billion of those initial funds have been spent so far, demanding accountability at an upcoming congressional briefing.

Microsoft unpacks last week’s outage. 

Microsoft has traced last week's nearly 24-hour global Microsoft 365 outage to a latent Windows defect in its cloud networking gateway servers. According to Redmond's preliminary incident report, the bug, introduced via a server-specific fix deployed back in July, caused severe "lock contention," where competing internal processes exhausted processor and memory capacity on newer, multi-core hardware.

The defect remained hidden for over three months until a shift in Microsoft Teams traffic collided with a 14 percent reduction in server capacity. As servers crashed, automated failover systems rerouted connections to healthy endpoints, triggering a cascading failure that rendered over 80 percent of servers unresponsive and severely disrupted Exchange Online.

Engineers eventually mitigated the crisis by rolling back the faulty connection-handling update and gradually restoring traffic. While Microsoft confirms customer Windows devices were unaffected and require no configuration changes, the incident underscores how automated recovery mechanisms can inadvertently amplify systemic cloud failures.

Commerce Secretary Lutnick defends less stringent export controls on advanced AI chips. 

According to The New York Times, Commerce Secretary Howard Lutnick has been advocating for less stringent export controls on advanced artificial intelligence chips, rules that directly affect Southeast Asian data centers with financial ties to his former firm, Cantor Fitzgerald.

While the US bans the direct sale of cutting-edge Nvidia chips to China, Chinese tech firms can legally rent computing power from facilities in countries like Malaysia, Indonesia, and Thailand. Lutnick, who transferred leadership and ownership of Cantor Fitzgerald to his sons, has pushed to limit new regulations to only seven countries, exempt existing data centers, and require government notifications rather than strict licenses.

Cantor Fitzgerald has arranged major financing for several of these data center operators, including Singapore-based Bitdeer, which plans to supply Nvidia computing power to Chinese tech giant Tencent. National security analysts and Democratic lawmakers have raised concerns over potential conflicts of interest and warned that narrower rules will allow Beijing to easily bypass restrictions.

Conversely, a Commerce Department spokeswoman stated that Lutnick bases his policy decisions strictly on US national interests, does not discuss department matters with his sons, and has complied with all federal ethics agreements. Furthermore, proponents of a lighter regulatory touch argue that overly broad restrictions could stifle American innovation and harm US companies like Nvidia.

Italy fines Meta 887 million dollars for privacy violations. 

Italy’s data protection authority has fined Meta 805 million euros, roughly 887 million US dollars, for accessing Android users’ photos, videos, and metadata without specific consent or proper notification.

According to the regulator, the violation stems from Facebook’s gallery suggestions feature, introduced in 2023. The app bundled camera roll access into a single, generic authorization prompt, leaving roughly 20 million Italian users unaware they were permitting Facebook to scan their personal galleries for post ideas. Regulators noted the infringement remains ongoing for users who have not updated their apps, and cited Meta’s lack of cooperation during the investigation when calculating the penalty.

Meta announced it will appeal, calling the fine disproportionate. Moving forward, Menlo Park must implement a separate, transparent consent prompt detailing how gallery data is accessed and how users can opt out.

Telegram’s new widgets may deliver a RAT.  

Kaspersky warns that a remote code execution vulnerability in Telegram’s newly launched widgets feature may have been exploited in the wild to deliver WebRAT, a Rust-based trojan also known as Salat Stealer. According to the Russian security firm, visiting compromised Telegram channels triggered a hidden PowerShell script to install the malware, which steals cryptocurrency wallets and Telegram account data.

The alert follows proof-of-concept videos from researchers at V12 demonstrating exploits against Telegram’s Android and Windows clients, which they attribute to running the QuickJS JavaScript engine without a sandbox.

While Kaspersky advises users against launching widgets until patched, Telegram CEO Pavel Durov dismissed the claims, stating users cannot be hacked merely by receiving a message and confirming that server-side patches have already resolved the underlying scenario.

The FBI warns of ARTEX pen testing abuse. 

The Federal Bureau of Investigation has issued a Flash notice warning that threat actors are increasingly weaponizing ARTEX, an open-source, artificial intelligence-driven penetration testing framework, to execute rapid, global cyber intrusions. Since at least August 2026, attackers have used the tool to target sectors including finance, government, defense, and healthcare, compromising over two thousand downstream tenants in a ConnectWise Cloud Backup supply chain breach and stealing data on roughly 67,000 individuals from South Korean financial institutions.

Released on GitHub in July 2026, ARTEX plugs into external large language models to automate reconnaissance and exploitation, enabling some intrusions to achieve privileged access in under three hours. According to the FBI, attackers predominantly pair the framework with Chinese-origin models like DeepSeek Flash due to low costs and minimal guardrails.

Although the creator removed the original repository in October following widespread abuse, community forks and commercial variants continue to proliferate. The Bureau notes that ARTEX’s speed leaves defenders little time to respond, urging organizations to bolster defenses on overlooked portals and third-party integrations.

Refrigeration controllers see opportunistic scanning. 

Threat actors began attempting to exploit newly disclosed vulnerabilities in Copeland’s XWEB EVO refrigeration controllers less than two hours after the flaws went public, according to security firm VulnCheck.

Researchers uncovered 14 flaws, each capable of yielding root privileges, after similar controllers were targeted in an October attack on 700 French supermarkets that France attributed to Iran. While nine vulnerabilities affect older firmware, five impact the latest builds, including predictable daily passwords derived from the device's exposed MAC address. Because Copeland ended support for the XWEB EVO line in March 2026, these flaws will not be patched.

VulnCheck’s sensors detected opportunistic scanning targeting unauthenticated flaws on November 12 and 13. With nearly 2,000 hosts exposed online, researchers urge operators to isolate these legacy devices behind a VPN and migrate to supported models.

Publicly available AI models dramatically accelerate vulnerability discovery. 

Cybersecurity firm Forescout Technologies reports that publicly available artificial intelligence models can dramatically accelerate vulnerability discovery in critical infrastructure software on a tight budget.

Using a custom AI-assisted research harness to examine 10 popular TCP/IP stacks, Forescout researchers compressed roughly nine months of work into just 90 hours, combining 10 hours of AI scanning with 80 hours of manual validation. The effort uncovered 99 flaws, including 66 zero-day vulnerabilities, at a static analysis cost of roughly $935, or $2.58 per finding. (Researchers noted binary analysis proved six times costlier and twice as slow as source-code review.)

Forescout VP of Research Daniel dos Santos warned that while AI still requires human expertise, it fundamentally alters vulnerability economics, allowing defenders and attackers alike to find embedded software bugs faster and cheaper.

Monday business briefing. 

Venture capital and mergers and acquisitions in the cybersecurity sector remain heavily focused on artificial intelligence, cloud security, and managed services, according to recent industry announcements.

On the funding front, Israeli AI detection and response platform Daylight led the pack with an $80 million Series B round backed by Goldman Sachs Alternatives. Fellow Israeli startups also saw major investments as Bay emerged from stealth with $50 million, cloud backup provider Dreeze raised $33 million, and SecOps platform Brava secured $15 million. Meanwhile, US firms Lazarus AI and WhiteMatter raised $43 million and $20 million respectively, and Italian automation startup CyberKord closed $3 million.

In M&A activity spanning four countries, compliance platform Vanta agreed to acquire Israeli agentic security firm Pillar Security, while Cogent acquired AI penetration testing startup Specular. Elsewhere, ShorePoint absorbed federal contracts from Coalfire Federal, while managed security and consulting consolidation saw acquisitions by Brite, ASI Solutions, Entech, and Saepio.

 

A refrigerator company learns it’s secretly leading the laundry revolution. 

Stirling Marathon, a modest appliance maker in Elora, Ontario, recently learned it was poised to revolutionize the North American laundry industry. There was just one minor catch: the company only makes refrigerators, freezers, and coolers.

According to CBC News, the sudden surge in fame was courtesy of an artificial intelligence scam. Fraudsters blended legitimate company marketing photos with AI-generated footage on YouTube to hawk a fictitious "bombshell" washing machine called "the Keel."

Customer service representative Jordan Cuddihy first realized something was amiss on a Wednesday morning after fielding puzzled inquiries from across the continent, and watching hopeful bargain-hunters physically show up at the facility. A quick Google search revealed the phantom appliance, prompting YouTube to eventually pull the plug on both the video and the channel. For Stirling Marathon, it was a brief, unsolicited brush with industrial disruption, proving that in the age of AI, you don’t actually need to manufacture a product to generate buzz.

And that’s the CyberWire.

For links to all of today’s stories, check out our Daily Briefing at the cyberwire dot 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 the show, please share a rating and review in your podcast app. Please also fill out the survey in the show notes or send an email to cyberwire@n2k.com

 

N2K’s lead producer is Liz Stokes. We’re mixed by  Tré Hester, with original music by and sound design Elliott Peltzman. Our contributing host is Maria Varmazis. Our executive producer is Jennifer Eiben. Peter Kilpe is our publisher. And I’m Dave Bittner. Thanks for listening.