Collaborate Disseminate
A Chinese-speaking threat actor is suspected to be behind a fresh wave of cyber attacks targeting government organizations mainly located in Central Asia, including Afghanistan, Kyrgyzstan, Tajikistan, Uzbekistan, Kazakhstan, and the Syrian Arab Republ… Continue reading Suspected Chinese-Speaking Hackers Target Central Asian Governments With OctLurk and SilkLurk
Microsoft has quietly pushed out a new OneDrive Photos app to Windows 11 users. What’s it all about? Continue reading Did a OneDrive Photos app just appear on your PC? Here’s what it does (and how to get rid of it)
alternative_right shares a report from Axios: Google has basically stopped sending people to websites (including our site) for answers and information. Instead, it’s using AI to answer them on its platform, in its words. Chartbeat data shared with Axio… Continue reading Publishers Are Losing Google Traffic As AI Answers Replace Links
If you have a PC enrolled in the Windows Insider Program, it’s highly likely that you have a new download today.
The post Microsoft Issues Five New Insider Builds appeared first on Thurrott.com.
Continue reading Microsoft Issues Five New Insider Builds
Happy Friday! We close out July with a delightfully short Ask Paul ahead of a weekend that can’t happen quickly enough.
The post Ask Paul: July 31 ⭐ appeared first on Thurrott.com.
Continue reading Ask Paul: July 31 ⭐
What you’ll get back depends on how much you spent – but don’t expect much. Continue reading Amazon is splitting its $600 million tariff refund with customers – here’s who’s eligible
The chart is interesting.
On the IPI benchmark, Opus 5 improved over Opus 4.8, reducing the probability of an attacker succeeding within 15 attempts from 5.5% to 2.0%, and from 0.5% to 0.2% on 1 attempt. It also improved on Sonnet 5 (5.9% at k=15) and Mythos 5 (2.6%), making it the most robust model evaluated. Opus 5 also outperformed all non-Claude models on this benchmark. The most robust non-Claude model was Muse Spark at 16.5% within 15 attempts—more than eight times Opus 5’s rate. The most capable GPT 5.6 variant, Sol, was comparable to its predecessor GPT 5.5 (20.0% versus 20.8% within 15 attempts), and was 10 times as likely to be successfully attacked as Claude Opus 5 at 2.0%. The other GPT 5.6 variants are less robust, at 30.4% (Terra) and 43.9% (Luna). A single attempt against GPT 5.6 Sol succeeded 3.1% of the time, higher than the 2.0% an attacker achieved against Opus 5 after fifteen attempts…
Continue reading Anthropic’s Opus 5 Is Better at Resisting Prompt Injection
From the Thomson Reuters Foundation, a welcome email describes the situation in South Africa and then turns to global support: I’m getting in touch to share some new reports and resources to support media freedom work in East and Southern Africa. Journ… Continue reading RESOURCE: Thomson Reuters Foundation provides free resources and legal help for independent media around the world
New York has sued prediction-market platform Kalshi, alleging it operates an “illegal gambling operation” without state authorization. “No matter what they call themselves, prediction markets like Kalshi are gambling platforms, plain and simple,” said … Continue reading New York Sues Kalshi For Running ‘Illegal Gambling Operation’
Disclosure: This article was created in collaboration with Status Labs. Continue reading How Status Labs Helps Brands Get Cited in ChatGPT: The Data Behind AI Search Visibility