“Drunk” AI is terrible at keeping secrets
AI models taught to write like drunk people became easier to jailbreak and more likely to leak secrets shared in confidence. That is the finding of UNSW Sydn...
Articles mapped to MITRE ATT&CK techniques. Select a technique to view matching articles.
370 articles found
AI models taught to write like drunk people became easier to jailbreak and more likely to leak secrets shared in confidence. That is the finding of UNSW Sydn...
Infostealer malware is increasingly becoming the bridge between a compromised developer workstation and an enterprise cloud environment, with Lumma, RedLine,...
A newly observed MacSync campaign shows a marked evolution in macOS-focused crimeware, replacing relatively simple AppleScript-driven delivery with layered b...
It was meant to make life simpler: a secret email address to which developers can send a message and create an issue in their GitLab project. But poor securi...
Two actions-cool GitHub Actions have been disabled for a second time after the repositories became accessible last week, months after they were compromised d...
A recent Verizon study found that vulnerability exploitation became the most common initial access vector, appearing in 31%…
SentinelOne has announced the expansion of Wayfinder Threat Hunting to the major public cloud services: AWS, Azure, and Google Cloud. It’s the latest offerin...
Business Email Compromise (BEC) operates on a familiar playbook. Threat actors breach a mailbox, silently monitor operations, map approval chains, and ultima...
The logistics sector has become the target of a new malicious cyber campaign that distributes an Android spyware codenamed Corp MDM. According to Have I Been...
AI coding agents are changing how quickly developers can build and ship software as well as how quickly credentials can become exposed. According to GitGuard...
ClickFix has become the most common way attackers get into enterprise networks, and it does it without an exploit, an attachment, or a file on disk. Our new ...
A malicious Firefox extension masquerading as a PDF identity-verification utility has been found targeting Google accounts through session-cookie theft and a...
Disruptive attacks on public-facing services, financially motivated cybercrime and compromises of shared technology providers are increasing cybersecurity ri...
A firewall is supposed to be the barrier between attackers and the enterprise network, but that barrier can itself become a threat actors’ tool. Check Point ...
Technology company F5 fixed a critical remote code execution vulnerability in its BIG-IP Access Policy Manager (APM) platform on Tuesday. The flaw impacts de...
The ShinyHunters cybercriminal organization on Tuesday replaced agency images on the FBIjobs.gov site with a photo of a Pokemon that has become the group’s d...
How many bugs have you missed because you didn’t send quite enough HTTP requests?
Microsoft has warned that the EvilTokens phishing-as-a-service platform has become a major driver of AI-enabled device-code phishing, compromising more than ...
A threat actor linked to the ongoing Graphalgo software supply-chain campaign has expanded beyond npm and PyPI, using malicious Terraform providers and Go mo...
Cisco Duo is the best MFA for most buyers comparing on price and speed published per-user tiers, a free small-team floor, and device trust included while Mic...