A Chinese startup, Sand AI, appears to be blocking certain politically sensitive images from its online video generation tool.
A China-based startup, Sand AI, has released an openly licensed, video-generating AI model that’s garnered praise from entrepreneurs like the founding director of Microsoft Research Asia, Kai-Fu Lee. But Sand AI appears to be censoring the hosted version of its model to block images that might raise the ire of Chinese regulators from the hosted version of the model, according to TechCrunch’s testing.
Earlier this week, Sand AI announced Magi-1, a model that generates videos by “autoregressively” predicting sequences of frames. The company claims the model can generate high-quality, controllable footage that captures physics more accurately than rival open models.
Microsoft security chief Charlie Bell says the SFI’s 28 objectives are “near completion” and that 11 others have made “significant progress.”
Microsoft, touting what it calls “the largest cybersecurity engineering project in history,” says it has moved every Microsoft Account and Entra ID token‑signing key into hardware security modules or Azure confidential VMs with automatic rotation, an overhaul meant to block the key‑theft tactic that fueled an embarrassing nation‑state breach at Redmond.
Just 18 months after rolling out a Secure Future Initiative in response to the hack and a scathing US government report that followed, Microsoft security chief Charlie Bell said five of the program’s 28 objectives are “near completion” and that 11 others have made “significant progress.”
In addition to the headline fix to put all Microsoft Account and Entra ID token‑signing keys in hardware security modules or Azure confidential virtual machines, Bell said more than 90 percent of Microsoft’s internal productivity accounts have moved to phishing‑resistant multi factor authentication and that 90 percent of first‑party identity tokens are validated through a newly hardened software‑development kit.
The firm has stopped taking orders on its website and apps, including for food and clothes.
Marks & Spencer (M&S) says it has stopped taking online orders as the company struggles to recover from a cyber attack.
Customers began reporting problems last weekend, and on Tuesday the retailer confirmed it was facing a "cyber incident".
Now, M&S has entirely paused orders on its website and apps - including for food deliveries and clothes - and says it will refund orders placed by customers on Friday.
The firm's shares fell by 5% following the announcement, before recovering.
Online orders remained paused on Saturday morning.
"We are truly sorry for this inconvenience," the retailer wrote in a post on X.
"Our experienced team - supported by leading cyber experts - is working extremely hard to restart online and app shopping.
"We are incredibly grateful to our customers, colleagues and partners for their understanding and support."
Bell Ambulance and Alabama Ophthalmology Associates have suffered data breaches affecting over 100,000 people after being targeted in ransomware attacks.
One of them is Milwaukee, WI-based Bell Ambulance, which provides ambulance services in the area. The company revealed last week in a data security notice that it detected a network intrusion on February 13, 2025.
An investigation showed that hackers gained access to files containing information such as name, date of birth, SSN, and driver’s license number, as well as financial, medical and health insurance information.
Bell Ambulance did not say in its public notice how many individuals are impacted, but the Department of Health and Human Services (HHS) data breach tracker revealed on Monday that 114,000 people are affected.
The Medusa ransomware group announced hacking Bell Ambulance in early March, claiming to have stolen more than 200 Gb of data from its systems.
The second healthcare organization to confirm a data breach impacting more than 100,000 people is Birmingham, AL-based ophthalmology practice Alabama Ophthalmology Associates.
SK Telecom, South Korea’s largest telecom company, disclosed a data leak involving a malware infection.
SK Telecom is South Korea’s largest wireless carrier — it has tens of millions of subscribers and holds roughly half of the local market.
The company revealed on Tuesday in a Korean-language statement posted on its website that it detected an intrusion on April 19. An investigation showed that the attackers deployed malware and managed to obtain personal information belonging to customers.
Following the incident, SK Telecom is offering customers a free SIM protection service designed to prevent SIM swapping, which suggests that the leaked data could be leveraged for such activities.
A detailed analysis of a multi-stage card skimming attack exploiting outdated Magento software and fake image files.
In today’s post we’re going to review a sophisticated, multi-stage carding attack on a Magento eCommerce website. This malware leveraged a fake gif image file, local browser sessionStorage data, and tampered with the website traffic using a malicious reverse-proxy server to facilitate the theft of credit card data, login details, cookies, and other sensitive data from the compromised website.
The client was experiencing some strange behaviour on their checkout page, including clients unable to input their card details normally, and orders not going through. They contacted us for assistance. Thinking this would be a straightforward case of credit card theft instead what we found was actually a fascinating and rather advanced malware which we will explore in detail in this post.
A new attack technique named Policy Puppetry can break the protections of major gen-AI models to produce harmful outputs.
A surveillance tool meant to keep tabs on employees is leaking millions of real-time screenshots onto the open web.
Your boss watching your screen isn't the end of the story. Everyone else might be watching, too. Researchers at Cybernews have uncovered a major privacy breach involving WorkComposer, a workplace surveillance app used by over 200,000 people across countless companies.
The app, designed to track productivity by logging activity and snapping regular screenshots of employees’ screens, left over 21 million images exposed in an unsecured Amazon S3 bucket, broadcasting how workers go about their day frame by frame.
M-Trends 2025 data is based on more than 450,000 hours of Mandiant Consulting investigations. The metrics are based on investigations of targeted attack activity conducted between Jan. 1, 2024 and Dec. 31, 2024. Key findings in M-Trends 2025 include:
55% of threat groups active in 2024 were financially motivated, which marks a steady increase, and 8% of threat groups were motivated by espionage.
Exploits continue to be the most common initial infection vector (33%), and for the first time stolen credentials rose to the second most common in 2024 (16%).
The top targeted industries include financial (17.4%), business and professional services (11.1%), high tech (10.6%), government (9.5%), and healthcare (9.3%).
Global median dwell time rose to 11 days from 10 days in 2023. Global median dwell time was 26 days when external entities notified, 5 days when adversaries notified (notably in ransomware cases), and 10 days when organizations discovered malicious activity internally.
M-Trends 2025 dives deep into the aforementioned infostealer, cloud, and unsecured data repository trends, and several other topics, including:
Democratic People's Republic of Korea deploying citizens as remote IT contractors, using false identities to generate revenue and fund national interests.
Iran-nexus threat actors ramping up cyber operations in 2024, notably targeting Israeli entities and using a variety of methods to improve intrusion success.
Attackers targeting cloud-based stores of centralized authority, such as single sign-on portals, to gain broad access.
Increased targeting of Web3 technologies such as cryptocurrencies and blockchains for theft, money laundering, and financing illicit activities.
We've previously, publicly and privately, analysed vulnerabilities in various ‘Backup and Replication’ platforms, including those offered by Veeam and NAKIVO - both of which have struggled to avoid scrutiny and in some cases, even opting to patch issues silently.
However, we’re glad to see that sense prevails - kudos to NAKIVO for acknowledging CVE-2024-48248 from our previous research and publicly responding to a new XXE vulnerability (CVE-2025-32406).
Backup and Replication solutions have become prime targets for ransomware operators for logical reasons — Veeam, for instance, has already seen widespread exploitation in the wild.
Combined with AI, polymorphic phishing emails have become highly sophisticated, creating more personalized and evasive messages that result in higher attack success rates.