notice..
> Delta is blocking AI agents from booking flights
> JPMorgan Chase and Charles Schwab block agents from accessing accounts
> Yelp requires paid data licensing for agents
> Amazon has blocked Grok bot and other shopping agents.
the internet is starting to block AI agents that were supposed to do everything for us
#OpenIntuneBaseline Windows v4.0 is here! 🎉
Highlights include:
* Continuing to be the most agile and cutting-edge Intune baseline on the planet!
* A shake-up of Compliance policies, allowing true grace period granularity (and ditching the EAS Password policy because it's 🗑️)
* Edge settings for days! Simple support for the Edge Management Service and security hardening and user experience improvements.
* Updated M365 Apps Security baseline alignment.
* Fixed MS breaking the in-box app removal policy.
* Defender behaviour tweaks based on real customer feedback, not arbitrary check-boxes.
* Deprecated and obsolete settings sent to live on a farm.
As always:
✅ Take what you want
✅ Change what doesn't suit your environment
✅ Test it before yeeting to 40,000 endpoints on Friday afternoon
❌ Don't treat any security baseline as a magical compliance button
Check out the full release notes at: https://t.co/FS3CVjDVxz
Then, head on over to the OIB Deployer to deploy the new and updated policies, or check your current config vs 4.0 with the Settings Validator!
https://deploy[.]https://t.co/1quHU5axSU
And, uhh, keep an eye on Friday, too ;)
#OIB #Windows #EndpointSecurity #Community #Capybara
The AI giant acknowledges that it could have done far more to prevent its AI agents from going rogue. But it still fails to explain why it didn't see this fiasco coming.
https://t.co/YGVGmAh1iG
He copied a value out of the Windows Event Log.
Pasted it into his special browser.
And he was signed into Microsoft Entra as the user who had just authenticated.
No password. No phishing link. No stolen private key.
@MGrafnetter's Black Hat USA 2026 research...
This is another important distinction between on-prem and the cloud.
With on-prem, Microsoft issues a CVE and now its up to you to find out if your tenant was exploited and clean up the mess.
When something similar happens on Entra ID, Microsoft will put in some reasonable effort to find out which customers were impacted and then informing them. It doesn't just stop there, they help customers fix them to as needed.
Case in point, here's a great example from the nAuth abuse for Entra ID and Microsoft's response 👇
If you found this interesting, register for @wpninjasus in Jan and join all the fun with @techspence and me digging into this and more in our showdown 😂
https://t.co/JWBL2cPXEx
Plus there is an awesome cast of folks who will be there
🍿🧨💥
Super impressed with @huggingface's breakdown of the AI agent autonomous cyber attack from OpenAI, including technical timeline, & interactive replay (AND how they defended against the attack)!
Here is the interactive attack replay vid and source:
https://t.co/SVsc5ClRcW
Starting today, OpenAI's GPT-5.6 family of models is available in Microsoft 365 Copilot—in Word, Excel, PowerPoint, Chat and Cowork. Learn more: https://t.co/NGK3Q55BZ6
Highly recommend Passkey Managed Saas Alerts or Huntress ITDR for this. One of the only current ways to protect against this. #Passkey#PasskeyTechnology#MSP#ITSP https://t.co/2rBJ6u0TOV