Google on Monday revealed that adversarial use of artificial intelligence (AI) by threat actors has gone beyond manual prompting to deploying autonomous multi-agent offensive frameworks like ARTEX, CAIRN, CyberStrikeAI, and PentAGI to execute multi-stage intrusions at greater operational velocity.
In tandem, threat actors are leveraging the power and capabilities of large language models (LLMs)
A default password opened doors it never should have. A trusted update delivered malware. Even a security feature became a way to break in. Different stories, same problem: the things meant to protect us aren't always doing their job.
Meanwhile, AI is making bug hunting faster, old flaws are still paying off, and attackers keep finding easy ways in. Some attacks take real skill. Others barely
Six flaws in Danfoss AK-SM800A refrigeration controllers could let a remote attacker take over a controller without logging in, Dutch security consultancy Midnight Blue said earlier this week.
No patch is available, and the researchers say the newest firmware, version 5.4.0, does not fix the flaws.
Government cybersecurity agencies in Belgium, Ireland and Germany told operators this week to
Cybersecurity researchers have disclosed details of a new variant of CoffeeLoader, a malware designed to deliver additional payloads to infected machines.
"On launch, the NSIS-based installer executable dropped a file named '7z.dll' into a temporary directory and ran it through rundll32.exe," Kaspersky said in an analysis published today.
"The '7z.dll' file was a patched version of the
A financially motivated threat activity cluster tracked as REF3364 has been attributed to a Linux-focused cryptocurrency mining operation that tampers with the readings used to detect high GPU workloads.
Central to the intrusion is a toolkit called SANSAR, a Rust-based orchestrator that receives commands over the TOR anonymity network to manage CPU and GPU miners. The activity was observed in
Attackers' AI agents have been hunting for weak spots in internet-exposed AI apps such as LiteLLM, One API, and Dify to steal the AI provider keys stored inside, according to the agents' own records.
The records, which Amazon's threat intelligence team captured with decoy servers, describe stolen keys being used to run the AI models that powered the search for more keys.
The team said this
An investigation into suspicious activity within a Russian software developer's network has uncovered a months-long compromise of its development infrastructure and a software code-signing server.
The activity, documented by Kaspersky's ICS CERT team after it was contacted by the affected developer in June 2025, has been attributed to a threat actor named BO Team (aka Black Owl, Hoody Hyena,
Meta was fined β¬805 million in Italy yesterday over a Facebook for Android feature that suggests photos and videos from users' phone galleries for them to post.
The Garante, Italy's data protection authority, found that the feature was enabled by default and that the app accessed Italian users' photos, videos, and related metadata without their explicit consent.
The company has 60 days to
The compromise of Anthropic's official Claude Code extension on Cursor's extension marketplace was the result of the theft of an access key that was exposed in the wake of a breach targeting Fireworks AI.
"An attacker obtained an access key we had issued to Fireworks," Cursor said in a Friday update. "The key was exposed by Fireworks, and the attacker used it to copy data from a read-only
The U.S. Cybersecurity and Infrastructure Security Agency (CISA) on Friday added a high-severity security flaw impacting UnrealIRCd, an open-source Internet Relay Chat (IRC) server, to its Known Exploited Vulnerabilities (KEV) catalog, following reports of active exploitation.
The vulnerability, tracked as CVE-2026-114713 (CVSS score: 8.6), is a case of improper neutralization of delimiters