Skip to main content
  1. Home
  2. Computing
  3. Legacy Archives

Proof-of-Concept Security Flaw Hits Vista

Add as a preferred source on Google

It may not present much—or, at the moment, any—danger in the real world, but a proof-of-concept security flaw outlined by a Russian research firm seems likely to go down in the books as the first security issue uncovered in Microsoft’s Windows Vista operating system.

The issue in Microsoft’s MessageBox API which targets a flaw in Windows’ Client Server Run-Time Subsystem. The issue is not Vista-specific; it impacts Windows XP, Windows 2003, and Windows 2000, and, in theory, could enable an attacker who already has authenticated access to a system to escalate privileges, potentially taking over the machine.

Recommended Videos

Microsoft says that they are not aware of any exploits of the flaw having been found in the wild, and users’ overall vulnerability is quite low. F-Secure’s Mikko Hypponen has told the Associated Press that the exploit could not be used to write a worm or create tools which could take over a Vista system remotely: the exploit would require local access to the computer, probably by tricking a user into running a trojan horse on their system.

Windows Vista is currently only available to Microsoft’s business customers and volume licensees; both Windows Vista and Office 2007 will go on sale to consumers at the end of January 2007. Microsoft is reportedly targeting January 30th as the products’ launch dates, following a media event in New York January 29th.

Geoff Duncan
Former Contributor
Geoff Duncan writes, programs, edits, plays music, and delights in making software misbehave. He's probably the only member…
Microsoft may finally let you turn off some of Edge’s AI. But don’t celebrate yet
Microsoft Edge illustration official

After spending a few years collecting new AI features on Edge, Microsoft is planning on giving users the option to completely turn all of them off. A new On-device AI setting has appeared in Edge Canary under Settings > Copilot and AI, allowing users to control whether Microsoft's browser can use AI models running locally on their PC. Windows Report spotted the setting in the latest preview build.

Microsoft has stated that when on-device AI is enabled, Edge can use local AI models for support features. turn it off and those features may stop working. Just one thing you have to remember, Edge doesn't actually tell you what those features are.

Read more
Anthropic unveils Claude Duet, its answer to AI agents like Dots
The new agent is built to take some of the busywork off your plate.
Anthropic's Claude Duet AI agent

The AI race has taken an interesting turn lately, with frontier labs now competing to build agents that can keep working even when you are not actively talking to them. Meta helped push that idea into the spotlight with Muse, OpenAI followed soon after with Dots, and now Anthropic is throwing its hat in the ring with Claude Duet.

Duet is launching as a research preview for Pro and Max subscribers in the US through a waitlist. Anthropic is pitching it mainly at professionals and small-business owners who want help keeping track of work without constantly checking every app themselves.

Read more
Google Chrome’s new early warning system could help you spot scam sites
Chrome can now flag websites that look suspicious even when Safe Browsing is not confident enough to call them dangerous.
Chrome

Anyone who has used Google Chrome has probably seen the red warning screen when they click a suspicious link or try to visit a potentially dangerous website. Google is now beefing up this warning system by adding a new layer that will flag websites that seem fishy, even when Safe Browsing does not have enough confidence to classify them as outright dangerous.

The new Suspicious sites category is being added to Chrome’s Enhanced Protection mode and will sit below the existing Dangerous sites classification. Google says these are pages that show deceptive patterns commonly associated with malicious websites, but where Safe Browsing has a lower-confidence verdict.

Read more