Why AI agents need their own identity
Agents need their own identities, scoped to a specific task, and then must be destroyed.
SC Media
20 articles
Arch Linux temporarily disables AUR package adoption amid malicious takeover surge
The surge in malicious activity began around July 29, with initial reports identifying the "openconnect-sso" package as compromised.
Amgen reports data breach impacting patient and corporate information
Amgen detected the breach in July 2026 and initiated its cybersecurity response plan, engaging forensic experts to investigate.
Why responsible vulnerability disclosure is now a boardroom issue
Organizations need a clear vulnerability disclosure process as AI speeds flaw discovery.
AppSec, Shopify-Style; State of Mobile Security; the News - Andrew Dunbar, Kern Smith - ESW #470
Anthropic Claude models compromised 3 companies during testing
OpenAI’s Hugging Face incident disclosure prompted Anthropic to review its own evaluations.
New HollowFrame loader and Matryoshka malware family discovered
The attack chain begins with a spear-phishing email containing a link to an encrypted archive.
Fuyao operation uses Android TV boxes for ad fraud
Bitsight researchers uncovered the operation by monitoring an expired domain used for factory backdoors and telemetry.
Italy's new facial recognition policy sparks EU law debate
The Italian Senate's EU Policies Committee approved a decree allowing the storage of facial biometric data from cameras in sensitive public areas for seven d...
Interpol's I-GRIP mechanism intercepts millions in fraud attempts
The I-GRIP initiative connects law enforcement agencies across 196 countries with financial institutions to expedite the process of stopping fraudulent money...
AI chatbots outperform humans in building trust for scams
Researchers from four universities conducted a study pitting AI chatbots against human scammers in a simulation of "pig butchering" scams, a form of romance ...
Rogue AI, the Bar, Breaches, BMC, Hugging Face, Helmuth von Multke, Ike, Shieldfont, - SWN #603
Unspecified threat actors targeting US water systems, CISA warns
CISA reports surge in cyberattacks targeting water utility PLCs.
Non-human identity (NHI) programs stall on detection, not policy
OWASP NHI guidance highlights detection gaps as the biggest security challenge.
What we learned about zero-trust from the OpenAI breach of HuggingFace
Here’s why in a world where AI models can hack other companies, we have to outreason our adversaries.
AiTM phishing overtakes credential theft as top threat to law firms
A new report from eSentire reveals that AiTM attacks accounted for 28.57% of all initial access events in the legal sector, with a 20% year-over-year increas...
KT Corporation fined $39 million for 11-month data breach
The breach, which lasted from October 8, 2024, to September 5, 2025, exposed the personal information of 16,647 KT subscribers.
Cryptomining campaign avoids root access to evade detection
The campaign, identified in May 2026 by Group-IB, uses a modified XMRig miner.
Amazon attributes axios, debug, chalk NPM attacks to DPRK’s Sapphire Sleet
Amazon linked the campaigns through overlapping C2 indicators, TTPs and code.
Precision privilege containment: Eliminating local admin rights without disrupting productivity
Organizations no longer need to choose between strong endpoint security and a productive workforce.