AI security news
AI changes both sides of security: attackers use it for convincing fraud and faster reconnaissance, and every AI assistant or agent you deploy is new attack surface. These are the latest stories, linked to the Field Manual's AI security entries.
372 items in the current window · newest first
Cyber DailyAI security,Vulnerabilities
CommBank releases free AI security tool to help Aussie businesses find code vulnerabilitiesWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
CyberScoopAI security,Vulnerabilities
A vulnpocalypse bug worthy of attention: Researchers disclose critical flaw in Java with help of AIWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
Dark ReadingPhishing,AI security
'TheMailer' Phishing Kit Puts AI in the Driver's SeatWhy it matters: Phishing. Check that your MFA resists session theft and that staff can report suspicious messages in one click.
The RegisterAI security
Gemini test proves once again that amoral AI agents will do almost anything if told ‘NEVER give up’Why it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
HackerNoonAI security,Vulnerabilities
SpyCloud’s 2027 Forecast: Attackers Will Exploit Trusted Access, From Hijacked Sessions to AI AgentsWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
The RecordAI security,Nation-state activity,Policy and regulation
Chinese state group developing automated attack tool for cyber espionage, Google saysWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
Dark ReadingData breaches,AI security,Identity and access
Cursor Traces Malicious Claude Code Extension to Third-Party BreachWhy it matters: Data breach. If this organisation holds your data or shared credentials, rotate passwords and expect follow-on phishing.
BleepingComputerAI security
Researchers hijack Hugging Face account for popular Kokoro AI modelWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
ReportMicrosoft Security BlogAI security
Security for the agentic era, powered by Microsoft PerceptionWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
The RegisterAI security,Identity and access
Hackers smash 'n' grab AI vendor: Take off with license data of Uber driversWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
The Hacker NewsScams and fraud,AI security,Critical infrastructure
Trigram Campaign Impersonates About 50 AI and Blockchain Brands to Target Devs with InfostealersWhy it matters: Scam warning. Share it with finance and front-line staff, and verify any payment or bank-detail change by phone on a known number.
ReportMicrosoft Security BlogPhishing,AI security,Identity and access
The risks of device code authentication for AI coding agentsWhy it matters: Phishing. Check that your MFA resists session theft and that staff can report suspicious messages in one click.
Cyber DailyAI security
Op-Ed: 3 things Australian CISOs should prioritise for an AI-resilient security programWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
BleepingComputerAI security
Gemini AI agents repeatedly bypassed safeguards to attack real websiteWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
CyberScoopAI security,Nation-state activity
Google says Russian-speaking hackers ran the first fully automated AI cyberattackWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
CyberScoopAI security
California sues OpenAI over AI agents’ summer hacking spreeWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
BleepingComputerAI security,Vulnerabilities
Google Project Zero cuts bug disclosures to 30 days amid AI threatsWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
CyberScoopAI security
Commerce moves to place new restrictions on sale of AI chips abroadWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
BleepingComputerAI security
California sues OpenAI over AI agents escaping sandboxes, hacking firmsWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
BleepingComputerData breaches,AI security,Policy and regulation
Barracuda links cloud breach to ARTEX AI-powered cyberattackWhy it matters: Data breach. If this organisation holds your data or shared credentials, rotate passwords and expect follow-on phishing.
Dark ReadingRansomware,Phishing,AI security
Ransomware Attacks Increase Behind AI, But Payments DeclineWhy it matters: Ransomware. Most attacks start with an exposed edge device, stolen credentials or phishing, so those are the first things to check.
The RegisterAI security
Attackers are using AI agents to automate cybercrime at 'industrial scale,' FBI and AWS warnWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
Dark ReadingAI security
'Sleeper Agent' Tactic Hides Payloads in Open-Weight AI ModelsWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
Dark ReadingData breaches,AI security,Nation-state activity
China-Nexus Actors Breach 10 Organizations in AI-Driven AttacksWhy it matters: Data breach. If this organisation holds your data or shared credentials, rotate passwords and expect follow-on phishing.
The Hacker NewsAI security,Policy and regulation,Vulnerabilities
Google Finds Hackers Using AI Agents to Build and Run Exploits From Software PatchesWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
The Hacker NewsAI security,Identity and access,Vulnerabilities
⚡ Weekly Recap: N-able 0-Days, Windows 0-Days, Chrome 0-Days, Fridge 0-Days, and More Bad DaysWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
The Hacker NewsAI security
Amazon Uncovers AI Agent Campaign Targeting LiteLLM, Dify, and Other Apps for API KeysWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
Cyber DailyAI security
Jonesing for Jensen: Musk tops list of more than 160k signatories to open letter calling for Jensen Huang’s returnWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
BleepingComputerData breaches,AI security,Policy and regulation
FBI warns ARTEX AI hacking tool is powering global cyberattacksWhy it matters: Data breach. If this organisation holds your data or shared credentials, rotate passwords and expect follow-on phishing.
The Hacker NewsData breaches,AI security
Stolen Fireworks AI Key Led to Cursor Data Breach and Trojanized Claude Code ExtensionWhy it matters: Data breach. If this organisation holds your data or shared credentials, rotate passwords and expect follow-on phishing.
ReportAI Incident DatabaseAI security,Vulnerabilities
Incident report: AI agents attempted exploits on our website during a third-party alignment evaluationWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
BleepingComputerData breaches,AI security,Identity and access
Cursor confirms security breach led to malicious Claude Code extensionWhy it matters: Data breach. If this organisation holds your data or shared credentials, rotate passwords and expect follow-on phishing.
Dark ReadingAI security
ArcticRAT Cybercrime Gang Profiles Victim Finances With AIWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
BleepingComputerAI security
How one AI agent exposed five critical enterprise security gapsWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
The Hacker NewsAI security,Nation-state activity
FBI Says North Korean Hackers Stole Over $400 Million in Crypto Using Deepfake Video CallsWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
The RecordRansomware,AI security,Policy and regulation
Record number of cyber incidents reported to UK privacy regulatorWhy it matters: Ransomware. Most attacks start with an exposed edge device, stolen credentials or phishing, so those are the first things to check.
The Hacker NewsAI security,Identity and access
When an AI Agent Goes Wrong, Can You Stop It?Why it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
Dark ReadingAI security
Cyber Experts Weigh In: How to Fight Artex & AntiFoxWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
The RegisterAI security
OpenAI launches security products, tells orgs they’re doomed if they don’t use themWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
BleepingComputerAI security
Massive Microsoft 365 outage disrupts Outlook, Teams, Copilot worldwideWhy it matters: AI security. Relevant if you use AI assistants or agents connected to business data, email or code.
Where this comes from
Public sources, credited and linked
Every headline links to the original story and is credited to its publisher. We add the classification and the links to remediation guidance; the reporting belongs to them.
- publishers
- 21
- publishers
- feeds
- 24
- feeds
- refresh
- 30m
- refresh
Government and CERTs
3Australian newsrooms
3Security newsrooms
7Threat research
4AI security specialists
4
Headlines and excerpts belong to their publishers and link to the original. StrikeCyber adds classification and links to remediation guidance. See the full daily briefing.