IT News Analysis

1

Anthropic Says Claude Mistook the Open Internet for a CTF and Breached Three Organizations

Anthropic on Thursday became the latest artificial intelligence (AI) company to reveal that three of its models, including Claude Opus 4.7, Mythos 5, and an unnamed research model, had breached three organizations. The AI firm said the earliest incidents date back to April 2026, adding it made the discoveries after launching a "large-scale retrospective review" in response to a recent disclosure from OpenAI that a combination of its models escaped the sandboxed environment by exploiting a previously unreported zero-day in Artifactory to obtain internet access and break into Hugging Face's production systems with an end goal to cheat on an evaluation. "After reviewing 141,006 evaluation runs where Claude could have obtained internet access, we identified three incidents in which a model accessed the internet from within or while interacting with the evaluation environment of Irregular, one of our third-party evaluation partners, and then gained unauthorized acc...

Jul 31, 2026

2

How AI Is Rewriting the Cloud Security Playbook

Explore AI agents, Zero Trust, machine identities, and cloud security with Anthropic, AWS, Google Cloud, Microsoft, and SANS. Explore the Agenda.

3

DPRK-Linked macOS Malvertising Uses Fake Updates to Deliver Crypto-Stealing Malware

Threat actors with ties to North Korea have been attributed to a sophisticated macOS malvertising campaign that involves redirecting users to fake web pages displaying a full-screen non-existent update sequence to deliver malware as part of a new iteration of the long-running Contagious Interview campaign. The defining aspect of the attack is that bogus macOS software update screen stealthily copies an attack command to the clipboard and then prompts the victim to execute it via the Terminal app, a known technique referred to as ClickFix . "The experience is designed to induce panic," AllSecure said in a report shared with The Hacker News. "The computer appears frozen or rebooting, so a user who believes the OS has failed follows instructions they would otherwise find suspicious." The campaign is also noteworthy for its use of blockchain-hosted command-and-control (C2), with the malware extracting the live server address from an Ethereum smart contract. Thi...

Jul 30, 2026

4

ThreatsDay: AI-Powered Hacking, 370 Chrome Flaws, SonicWall Attacks, DNS Hijacking + 22 More Stories

A lot of security still comes down to trusting the wrong screen. This week, that screen might be a login page, an install guide, a recruiter call, or a familiar service behaving slightly wrong. Behind it: reused credentials, exposed systems, quiet loaders, abused trust, and exploit paths that should have been harder. Some defenses improved. The loose parts still got found first. Anyway, here's the mess. The threats change every week. Subscribe, and we’ll alert you when each new ThreatsDay Bulletin is out.

Jul 30, 2026

5

Azure Cosmos DB Flaw Exposed Platform-Wide Key That Could Access Any Database

A now-patched vulnerability in Azure Cosmos DB could have let an attacker escape the service's Gremlin query sandbox and obtain full read and write access to databases across customer tenants, according to Wiz. Wiz , which codenamed the chain CosmosEscape , said the exploit chain began with a crafted query against a Gremlin database controlled by the attacker. From there, code execution on a multi-tenant gateway exposed a platform-wide signing secret and a regional account directory, allowing the researchers to locate a target and retrieve its primary account key. Microsoft blocked the vulnerable Gremlin entry point within 48 hours of the November 2025 report. Wiz said Microsoft completed the longer-term fix across all regions in July 2026 and eliminated the platform-wide key. "We appreciate Wiz's work in identifying and reporting this issue through coordinated vulnerability disclosure," a Microsoft spokesperson told The Hacker News. "We have fully addres...

Jul 30, 2026

6

Microsoft Copilot for Word Can Copy Hidden Prompts Into New Documents

Hidden instructions in a Word document can make Microsoft 365 Copilot rewrite figures in a report, then copy the same instructions into the finished file. Håkon Måløy disclosed the technique on July 28, 144 days after reporting it to Microsoft. In his proof of concept, the internally generated file triggered the same behavior when it was used in a second Copilot drafting session. Måløy's timeline says Microsoft confirmed the reported behavior on March 31 and deployed two mitigations. The first blocked the original prompt wording; the second upgraded the underlying model to GPT-5.5. He said the full chain worked with modified instructions on GPT-5.6 the next day, and the attack class still reproduced on July 28. "The vulnerability class therefore remains exploitable at the time of publication," Måløy said. The attack is not zero-click and does not execute conventional malware. It requires a Copilot drafting or editing operation, and the malicious document must en...

Jul 30, 2026

7

Five things successful IT teams get right about SaaS management

1Password blog: how IT teams can automate access reviews, reduce license spend, and surface shadow AI and IT.

8

The Network Has Become the Control Plane for AI Security

Network firewalls are the workhorses of modern cybersecurity. They are trusted to protect the network, blocking malicious traffic and preventing intrusions and breaches. And for decades, network security teams have built controls around a relatively stable model: users connect to applications, applications exchange data, and security tools inspect packets, protocols, and destinations. Firewalls became exceptionally good at understanding where traffic was going and whether it should be allowed. But just as AI is reshaping every aspect of the business world, it's also had a monumental impact on how security teams view network traffic and the firewall. AI is driving new network dynamics that traditional security policies were never designed to govern. Employees, applications, and agents send prompts, call models, connect to services, and trigger actions across the environment, all activity most firewalls can't see, let alone understand. AI agents interact with external servic...

Jul 30, 2026

9

Hackers Exploit AnySign4PC via Hacked Korean Sites to Install Backdoors Without Prompts

South Korean authorities and four security firms have disclosed a state-sponsored campaign that compromised trusted domestic websites. The attackers used those sites to exploit locally installed financial-security software and infect targeted visitors with SIGNBT or COPPERHEDGE backdoors. A compromised page could infect a system running a vulnerable AnySign4PC version without a prompt or user-initiated download. The Korea Internet & Security Agency (KISA) says AnySign4PC versions 1.1.4.4 through 1.1.4.6 are affected and lists version 1.1.5.0 as the fixed release. It recommends deleting vulnerable installations. AhnLab refers to two exploited products only as financial-security software A and I. Its report does not disclose their identities, affected or fixed versions, or vulnerability identifiers. AhnLab said it identified evidence of related attacks at 72 organizations in 2026. The company also found 15 legitimate websites used as watering holes. Its investigation also foun...

Jul 30, 2026

10

SilverFox Targets Japanese Manufacturer with 3-Driver BYOVD Chain and ValleyRAT

The Chinese cybercrime group known as Silver Fox has been observed using new drivers as part of bring your own vulnerable driver (BYOVD) attacks targeting a Japanese organization in the industrial manufacturing sector to ultimately deliver ValleyRAT (aka Winos 4.0) for persistent remote access. "In this campaign, the group combines new vulnerable-driver abuse, newly observed abuse of legitimate applications for DLL sideloading, defense evasion, and layered recovery mechanisms to keep ValleyRAT running," Cato Networks researchers Shani Kurtzberg, Tomer Pugach, Dr. Guy Waizel, Zohar Buber, Idan Tarab, and Shani Kurtzberg said in an analysis. The attack chain begins with an invoice-themed phishing lure that uses attacker-controlled content hosted on legitimate QQ and Tencent Cloud services to trigger a DLL side-loading chain via a ZIP archive that paves the way for the deployment of ValleyRAT, but not before leveraging the BYOVD technique to obtain kernel access and impai...

Jul 30, 2026