Microsoft has just announced a whole slew of new "AI" features for Windows, and this time, they'll be living in your taskbar.
Microsoft is trying to transform Windows into a "canvas for AI," with new AI agents integrated into the Windows 11 taskbar. These new taskbar capabilities are designed to make AI agents feel like an assistant in Windows that can go off and control your PC and do tasks for you at the click of a button. It's part of a broader overhaul of Windows to turn the operating system into an "agentic OS."
[...]
Microsoft is integrating a variety of AI agents directly into the Windows 11 taskbar, including its own Microsoft 365 Copilot and third-party options. "This integration isn't just about adding agents; it's about making them part of the OS experience," says Windows chief Pavan Davuluri.
↫ Tom Warren at The VergeThese "AI" agents will control your computer, applications, and files for you, which may make some of you a little apprehensive, and for good reason. "AI" tools don't have a great track record when it comes to privacy – Windows Recall comes to mind – and as such, Microsoft claims this time, it'll be different. These new "AI" agents will run in what are essentially dedicated Windows accounts acting as sandboxes, to ensure they can only access certain resources.
While I find the addition of these "AI" tools to Windows insufferable and dumb, I'm at least glad Microsoft is taking privacy and security seriously this time, and I doubt Microsoft would repeat the same mistakes they made with the entirely botched rollout of Windows Recall. in addition, after the Cloudstrike fiasco, Microsoft made clear commitments to improve its security practices, which further adds to the confidence we should all have these new "AI" tools are safe, secure, and private.
But wait, what's this?
Additionally, agentic AI applications introduce novel security risks, such as cross-prompt injection (XPIA), where malicious content embedded in UI elements or documents can override agent instructions, leading to unintended actions like data exfiltration or malware installation.
↫ Microsoft support document about the new "AI" featuresMicrosoft's new "AI" features can go out and install malware without your consent, because these features possess the access and privileges to do so. The mere idea that some application – which is essentially what these "AI" features really are – can go out onto the web and download and install whatever it wants, including malware, "on your behalf", in the background, is so utterly dystopian to me I just can't imagine any serious developer looking at this and thinking "yeah, ship it".
I'm living in an insane asylum.
More details from the Microsoft link:
We recommend that you only enable this feature if you understand the security implications outlined on this page. This setting can only be enabled by an administrator user of the device and once enabled, it's enabled for all users on the device including other administrators and standard users.
[...] Agentic AI has powerful capabilities today—for example, it can complete many complex tasks in response to user prompts, transforming how users interact with their PCs. As these capabilities are introduced, AI models still face functional limitations in terms of how they behave and occasionally may hallucinate and produce unexpected outputs. Additionally, agentic AI applications introduce novel security risks, such as cross-prompt injection (XPIA), where malicious content embedded in UI elements or documents can override agent instructions, leading to unintended actions like data exfiltration or malware installation.
Related: SUSE to Include Agentic AI in SLE 16
« "Walk My AI Walk", a #1 AI Country Song | First Large-Scale VR Use in the Classroom Extends the Boundaries of Education »
Related Stories
https://distrowatch.com/dwres.php?resource=showheadline&story=20085
SUSE has announced SUSE Linux Enterprise, which is schedule for release on November 4th, will be the first enterprise-focused Linux distribution to include agentic AI.
"SLES 16 introduces agentic AI, with an implementation of the Model Context Protocol (MCP) standard. The SUSE Linux agentic AI implementation gives enterprises a secure, extensible way to connect AI models with external tools and data sources, while preserving freedom to choose and extend their preferred AI providers without lock-in. It provides a resilient and secure foundation, combining long-term lifecycle guarantees and enterprise-grade automation."
SUSE has also stated SLE 16 will receive up to 16 years of support. Further details are provided in the company's announcement.
(Score: 4, Insightful) by Anonymous Coward on Saturday November 22, @03:30PM (4 children)
All you can do is roll your eyes over all of this nonsense until The Bubble pops.
(Score: 5, Informative) by Deep Blue on Saturday November 22, @04:48PM (2 children)
The current level of sanity (what ever it might be, seems kind of low anyways) won't be the same even after that.
How in the hell does MS think this is a good idea? They've gone full retard. Never go full retard.
(Score: 0) by Anonymous Coward on Saturday November 22, @05:18PM
"Thank you, Sir! May I have another?"
(Score: 5, Touché) by Gaaark on Saturday November 22, @08:48PM
They've BEEN full retard for a while. But people keep taking it.
--- Please remind me if I haven't been civil to you: I'm channeling MDC. I have always been here. ---Gaaark 2.0 --
(Score: 1, Interesting) by Anonymous Coward on Sunday November 23, @10:33AM
We can choose to use more robust technology.
But, I understand about communicating with other people, likely 'superiors' in employment echelons, that simply seem to have no idea about what's going on. They seem to be much like my cat, who thinks a nice warm road is an excellent place to take a nap. It simply cannot extrapolate a human's knowledge of the aftermath of an encounter with a car.
It seems no more serious than a bullet point on some status report. So we lost a reputation that took decades to build. Just change employer and start over. The big thing is CYA, so someone else gets blamed. Seems of little concern that bad things happen, the big thing is keeping yourself held harmless by delegation of responsibility.
I know one day, I will lose that cat because it did not realize the dangers of sleeping in the road, and one day I may lose my workplace as those in charge did not realize the value of their computing infrastructure.
I have to deal with laws of physics, where they deal with laws of men. My outcomes are determined by physical law, their outcomes are determined by psychology - and I am dependent on people to keep a job, so I often have to go along with things I consider absolute lunacy to stay employed.
(Score: 5, Informative) by liar on Saturday November 22, @03:39PM (2 children)
O&O ShutUp10 , is my friend. Get everything set up the way I like, then save (export) that setup. Next time I run Windows update, afterwards I run Shutup from Imported file and it tells me what Windows has reactivated with their update... and what it's turned back off.
https://www.oo-software.com/en/shutup10 [oo-software.com] "O&O ShutUp10++ means you have full control over which comfort functions under Windows 10 and Windows 11 you wish to use, and you decide when the passing on of your data goes too far. Using a very simple interface, you decide how Windows 10 and Windows 11 should respect your privacy by deciding which unwanted functions should be deactivated."
Noli nothis permittere te terere.
(Score: 3, Interesting) by ikanreed on Saturday November 22, @04:48PM
I remember when I first used macos in 2008 and found the only way to change some system settings was to install an app to do it.
I didn't realize I was just using the future of all computing platforms. I thought it was annoying horseshit for idiots. What a fool I was
(Score: 5, Insightful) by Gaaark on Saturday November 22, @08:50PM
Oh, man: "O&O ShutUp10".
Really?
I would so re-name it "Quiet, piggy!"
--- Please remind me if I haven't been civil to you: I'm channeling MDC. I have always been here. ---Gaaark 2.0 --
(Score: 3, Interesting) by Mojibake Tengu on Saturday November 22, @04:48PM (3 children)
Both GPT and Gemini are quite capable to write code in NASM. They even understand its macro language. Not sure about Copilot or Claude, though.
Still not the ways I can do, to match me at level, but they often perform much better than most of other humans...
Remember the Core Wars game?
Rust programming language offends both my Intelligence and my Spirit.
(Score: 5, Touché) by Deep Blue on Saturday November 22, @04:51PM (2 children)
"understand", really?
(Score: 3, Informative) by Mojibake Tengu on Sunday November 23, @12:24AM
Running code is the ultimate reality. It does not care about what people think about it.
It just executes.
In that sense, it becomes Philosophy. A Metaphysics, not Engineering anymore.
Rust programming language offends both my Intelligence and my Spirit.
(Score: 1, Interesting) by Anonymous Coward on Sunday November 23, @08:14AM
Four levels: https://www.youtube.com/watch?v=UGXkTmmkIFc [youtube.com]
(Score: 2) by istartedi on Saturday November 22, @05:16PM
Bonzi Buddy could have developed in to a complete OS. They were ahead of their time. Just like always, MS stealing from the real innovators.
Appended to the end of comments you post. Max: 120 chars.
(Score: 4, Informative) by FuzzyTheBear on Saturday November 22, @06:50PM
Where malware is a feature .. not a bug .. nuff said.
(Score: 4, Interesting) by VLM on Saturday November 22, @07:04PM (1 child)
The corporate responsibility issues must be insane.
So you signed a doc as an employee promising to NOT install malware as an acceptable use policy, malware was installed, "that was microsoft AI doing it you can't fire me nor even punish me".
(Score: 2) by corey on Monday November 24, @07:43PM
It’s hypocritical as well. Remember how Microsoft added the UAC annoyance whenever you had to do anything that required privileges like installing software? Then they removed it but it’s sort of still there. It’s because they view users as untrustworthy and stupid. But now this, they trust their AI LLM openly. Does it have to click through UAC warnings and approvals?
(Score: 4, Insightful) by Gaaark on Saturday November 22, @08:35PM
What they need to come up with is AI so my mother-in-law can say, "All i want is to get my Facebook back. Where did it go?" and have it bring it up for her.
Or "All i want to do is print this obituary" and have it bring up the print dialog, clear the print queue of the hundred times she's hit print, and then print the frigging obituary.
She does not need more shit to confuse her.
Seriously.
--- Please remind me if I haven't been civil to you: I'm channeling MDC. I have always been here. ---Gaaark 2.0 --
(Score: 4, Insightful) by jb on Sunday November 23, @07:17AM
Given that Ms Windows itself already ships with malware built in to it (hiding behind the treacherously redefined Newspeak term "telemetry"), why would any of its users care about it adding even more malware?
(Score: 2) by Rosco P. Coltrane on Sunday November 23, @01:15PM
and the best advertisement for Linux ever.
So Microsoft is essentially saying "We're pushing Copilot in every nook and cranny of the OS whether you like it or not, and it can give you viruses but it's really worth it!"
What a stupid company. 50 years of utter stupidity and the level of stupidity doesn't abate...