Skip to content

AI Cyber Defense, Windows & Silicon

OpenAI Teams With NATO on Cyber Defense as Microsoft Tests Remote Lock

OpenAI partnered with NATO's Communications and Information Agency on up to 300 billion monthly tokens for cybersecurity, Microsoft rolled out Remote Lock in Windows 11 Insider builds, and more than 400 leaders gathered at Semafor's Menlo Park conference.

By Lazy Tensor
Published · 4 min read

Restrained Economist-style conceptual print illustration on warm cream cotton paper showing a geometric shield protecting interconnected fiber-optic network nodes and a locked laptop silhouette
Frontier cybersecurity models and remote enterprise device management strengthen network resilience across allied institutions and corporate endpoints. (Illustration: Lazy Tensor) · Lazy Tensor Editorial Illustration
Your listening Queue & catch up
Unheard stories first, timed at 1×. Your selections stay first.

Add a story or build a catch-up queue to start listening.

Frontier cybersecurity partnerships, enterprise endpoint security controls, and executive debate over semiconductor leadership defined technology coverage on Tuesday as OpenAI expanded its artificial intelligence defense collaboration with NATO and European cyber agencies, Microsoft began testing Remote Lock for Windows 11 devices, and more than 400 leaders gathered in Menlo Park for Semafor's Silicon Valley & the World conference.

OpenAI Collaborates With NATO and Allied Agencies on AI Cyber Defense

OpenAI announced Tuesday that it is working with NATO's Communications and Information Agency (NCIA) to provide core allied institutions with advanced cybersecurity models through its Daybreak tool, with NATO consuming up to 300 billion raw input and output tokens per month to protect allied networks, according to OpenAI News. OpenAI also supported the 2026 Crossed Swords exercise organized by the NATO Cooperative Cyber Defence Centre of Excellence (CCDCOE), where operators trained alongside AI agents under human oversight. NCIA General Manager Dr. Dylan Browne and OpenAI Head of National Security Policy Sasha Baker highlighted how machine-speed vulnerability discovery strengthens digital infrastructure resilience.

Beyond NATO, OpenAI has provided its cybersecurity models and Daybreak tool to the Ministry of Digital Transformation of Ukraine, the European Union's cyber agency ENISA, and national cyber agencies in France (ANSSI), Germany, and Poland. ENISA used the models to identify and publish fixes for 4 vulnerabilities (1, 2, 3, and 4) in software used across EU institutions, while a November joint report from ENISA and CERT-EU reviewed 27 software projects using Daybreak and the Codex security plug-in, confirming that 95% of the 88 potential issues assessed by maintainers were genuine. Meanwhile, CERT Polska used OpenAI models to identify 6 vulnerabilities in MikroTik RouterOS and coordinated vendor fixes.

Microsoft Tests Intune Remote Lock in Windows 11 Insider Preview Builds

In enterprise endpoint management, Microsoft is testing a Remote Lock feature that allows IT administrators to lock lost or stolen Windows 11 devices without wiping user data, BleepingComputer reported. Shipping in Windows 11 Insider Preview builds in the Beta (Build 26320.10003) and Experimental (Build 26340.10013) channels, Remote Lock uses Microsoft Intune custom configuration profiles (under Windows 10 and later) to lock Microsoft Entra-joined, hybrid-joined, or MDM-enrolled devices running Windows Enterprise, Education, or Pro by setting LockState = 1 to lock and LockState = 0 to unlock.

Windows Insider communications lead Stephen Lines explained that Remote Lock signs out active sessions, disables PINs, passwords, local admins, and web sign-ins on the locked machine while keeping BitLocker enabled and allowing the user's credentials to work on other devices across restarts. Admins must keep BitLocker active since Safe Mode, USB booting, and the Windows Recovery Environment remain accessible unless separate Intune isolation or wipe controls are triggered. The preview builds also test individual passkey management on FIDO2 (Fast Identity Online 2) security keys following August privacy controls for microphone, camera, and location access, alongside 14-day read-only email scans for Microsoft 365 and Google Workspace.

Executives Weigh In on Jensen Huang at Semafor Conference in Menlo Park

At the two-day Silicon Valley & the World conference in Menlo Park hosted by Semafor cofounder Justin B. Smith, editor-in-chief Ben Smith, and tech editor Reed Albergotti, more than 400 CEOs, investors, and officials discussed former Nvidia CEO Jensen Huang, who had been listed as an event co-chair before his resignation last week, according to The San Francisco Standard. OpenAI CEO Sam Altman—who experienced his own 2023 ouster and reinstatement and whose company received a $30 billion investment from Nvidia—called Huang a personal hero, while Perplexity CEO Aravind Srinivas praised Huang's grit and said reinstatement is for the board to decide. Former White House AI czar David Sacks questioned reports involving Alice Huang and Secretary of State Marco Rubio, while hedge fund manager Bill Ackman called Huang an iconic CEO but cited legal uncertainty and a venture capitalist joked about avoiding The Battery in San Francisco.

Sources

Follow the coverage

Published reporting connected by Cybersecurity.

  1. OpenAI Releases ChatGPT Work Data as ARC Evals Tests Autonomous Agents
  2. Dreamteam Raises $53M as Anthropic Updates Claude and Amazon Builds
  3. Corma Hits $1B Value as Holiday Extras Uses AI & U.S. Caps Chips
  4. OpenAI Teams With NATO on Cyber Defense as Microsoft Tests Remote LockYou are reading this story

Comments (0)

No comments yet. Be the first to share a thought on this story.

The Lazy Tensor briefing

Get the next story in your inbox.

Independent reporting on the people, companies and ideas shaping AI and technology. Free, once a week.

More from Lazy Tensor