Vulnerability | Breaking Cybersecurity News | The Hacker News

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...

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.

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...

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...

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...