$ page 27/37

MCP security risks stem from AI architecture, not a patchable bug

Researchers warn MCP introduces architectural AI security risks that standard patches cannot fix, raising enterprise concerns over tool access.

2 min readMar 21

CISA orders agencies to patch exploited Cisco SD-WAN flaws

CISA has ordered federal agencies to patch actively exploited Cisco SD-WAN flaws that can hand attackers admin access to network infrastructure.

2 min readMar 21

ShinyHunters targets hundreds of websites in new Salesforce campaign

ShinyHunters’ claimed Salesforce Experience Cloud campaign shows how stolen identities and weak portal controls can expose data at scale.

8 min readMar 21

Aura confirms breach of 900,000 marketing contacts

Aura says an unauthorized party accessed nearly 900,000 marketing contact records, exposing names and email addresses and raising phishing risks.

8 min readMar 21

Leaked Ariomex database sheds light on Iranian crypto sanctions evasion

A leaked Ariomex database may reveal how Iranian actors used crypto to move funds under sanctions pressure, while exposing users to serious privacy ri

8 min readMar 21

Former defense contractor executive gets 7+ years for selling zero-days to Russia

A former L3Harris executive got 7+ years for selling stolen zero-days to a Russian broker, exposing insider risks in the cyber-industrial base.

9 min readMar 21

Google patches Chrome zero-day after confirming in-the-wild exploitation

Google issued an emergency Chrome update for CVE-2024-4671, a high-severity zero-day reportedly exploited in the wild.

2 min readMar 21

NCSC warns critical infrastructure firms to act now after severe attacks hit Polish energy providers

The NCSC is warning critical infrastructure operators to act now after severe disruptive cyber-attacks hit Polish energy providers.

2 min readMar 21

CISA flags actively exploited SolarWinds Web Help Desk flaw

CISA added critical SolarWinds Web Help Desk flaw CVE-2024-28986 to KEV, warning that attackers are already exploiting it in the wild.

2 min readMar 21

AI voice and virtual meeting fraud jumped 1210% in a year, Pindrop says

Pindrop says AI-powered voice and virtual meeting fraud surged 1210% in a year, raising risks for payments, call centers and remote work.

2 min readMar 21

Critical and high-severity n8n sandbox flaws open path to remote code execution

Two severe n8n sandbox flaws could enable remote code execution, putting self-hosted automation servers and stored secrets at risk.

2 min readMar 21

World Leaks claims 1.4TB Nike data breach as extortion-only attacks keep rising

Nike is investigating World Leaks’ claim of a 1.4TB data dump, highlighting the rise of extortion-focused attacks without confirmed encryption.

9 min readMar 21