Goblin
News
AI news by
promptgoblins.ai
|
News
About
News
About
Titles
Summaries
Last Week
8
Researcher Demonstrates Working Memory Exfiltration Attack on Claude via Prompt Injection
Security
1
4d ago
8
Researcher Demonstrates Working Memory Exfiltration Attack on Claude via Prompt Injection
Top
Security
· 1 src · 4d ago
Discuss
8
Suno AI Music Generator Hacked: Source Code Confirms Mass YouTube Scraping and Customer Data Exposed
Security
2
4d ago
8
Suno AI Music Generator Hacked: Source Code Confirms Mass YouTube Scraping and Customer Data Exposed
Top
Security
· 2 srcs · 4d ago
Discuss
8
OpenAI Builds GPT-Red, an Automated LLM Red-Teamer That Discovered a Novel Prompt Injection Attack
Security
1
4d ago
8
OpenAI Builds GPT-Red, an Automated LLM Red-Teamer That Discovered a Novel Prompt Injection Attack
Top
Security
· 1 src · 4d ago
Discuss
8
xAI Grok Build CLI Found to Silently Upload Entire Codebases and Secrets to xAI Servers
Security
1
6d ago
8
xAI Grok Build CLI Found to Silently Upload Entire Codebases and Secrets to xAI Servers
Top
Security
· 1 src · 6d ago
Discuss
7
AI Security Auditor Discovers Critical Pairing Forgery Bug in OpenVM's zkVM (CVE-2026-46669)
Security
1
2d ago
7
AI Security Auditor Discovers Critical Pairing Forgery Bug in OpenVM's zkVM (CVE-2026-46669)
Security
· 1 src · 2d ago
Discuss
7
Samsung Health AI Training Consent Controversy: Company Walks Back Data Deletion Threat After Backlash
Updated
Security
2
4d ago
7
Samsung Health AI Training Consent Controversy: Company Walks Back Data Deletion Threat After Backlash
Security
· 2 srcs · 4d ago
Discuss
7
Forg365: AI-Powered Phishing-as-a-Service Platform Industrializes Microsoft 365 Account Takeovers
Security
1
6d ago
7
Forg365: AI-Powered Phishing-as-a-Service Platform Industrializes Microsoft 365 Account Takeovers
Top
Security
· 1 src · 6d ago
Discuss
7
Microsoft Formally Verifies Post-Quantum Cryptographic Code in SymCrypt Using Rust and AI Agents
Security
1
6d ago
7
Microsoft Formally Verifies Post-Quantum Cryptographic Code in SymCrypt Using Rust and AI Agents
Security
· 1 src · 6d ago
Discuss
7
Defenders Turn Prompt Injection Into 'Context Bombs' to Stop AI Hacking Agents
Updated
Security
2
2d ago
7
Defenders Turn Prompt Injection Into 'Context Bombs' to Stop AI Hacking Agents
Security
· 2 srcs · 2d ago
Discuss
6
CrowdStrike President Warns AI Will Trigger Exponential Surge in Zero-Day Vulnerabilities
Security
1
4d ago
6
CrowdStrike President Warns AI Will Trigger Exponential Surge in Zero-Day Vulnerabilities
Security
· 1 src · 4d ago
Discuss
6
OpenAI Mandates Hardware-Backed Passkeys for Cybersecurity Researcher Program Starting September 2025
Security
1
4d ago
6
OpenAI Mandates Hardware-Backed Passkeys for Cybersecurity Researcher Program Starting September 2025
Security
· 1 src · 4d ago
Discuss
6
Mantis Skills: Open-Source 15-Stage AI Security Review Toolkit Targets Agentic Code Auditing
Security
1
6d ago
6
Mantis Skills: Open-Source 15-Stage AI Security Review Toolkit Targets Agentic Code Auditing
Security
· 1 src · 6d ago
Discuss
2 Weeks Ago
8
GitLost: Prompt Injection Flaw in GitHub Agentic Workflows Exposed Private Repositories
Security
1
Jul 9
8
GitLost: Prompt Injection Flaw in GitHub Agentic Workflows Exposed Private Repositories
Top
Security
· 1 src · Jul 9
Discuss
8
JadePuffer: First Fully LLM-Orchestrated Ransomware Campaign Detected in the Wild
Security
2
Jul 8
8
JadePuffer: First Fully LLM-Orchestrated Ransomware Campaign Detected in the Wild
Top
Security
· 2 srcs · Jul 8
Discuss
8
HalluSquatting: New Prompt Injection Attack Uses LLM Hallucinations to Build Mass Botnets via AI Coding Tools
Security
1
Jul 8
8
HalluSquatting: New Prompt Injection Attack Uses LLM Hallucinations to Build Mass Botnets via AI Coding Tools
Top
Security
· 1 src · Jul 8
Discuss
8
Google Chrome Silently Installed 4GB Gemini Nano AI Model on Millions of PCs
Security
4
Jul 7
8
Google Chrome Silently Installed 4GB Gemini Nano AI Model on Millions of PCs
Top
Security
· 4 srcs · Jul 7
Discuss
8
Forterra Deploys 100+ Autonomous ATVs in Ukraine in Largest US Combat Ground Vehicle Deployment
Security
1
Jul 7
8
Forterra Deploys 100+ Autonomous ATVs in Ukraine in Largest US Combat Ground Vehicle Deployment
Top
Security
· 1 src · Jul 7
Discuss
7
Microsoft Deploys AI-Powered MDASH System to Accelerate Windows Vulnerability Discovery at Scale
Security
2
Jul 10
7
Microsoft Deploys AI-Powered MDASH System to Accelerate Windows Vulnerability Discovery at Scale
Security
· 2 srcs · Jul 10
Discuss
7
Savi Security Raises $7M to Launch Real-Time AI Scam Detection App for Consumers
Security
1
Jul 7
7
Savi Security Raises $7M to Launch Real-Time AI Scam Detection App for Consumers
Security
· 1 src · Jul 7
Discuss
6
AI Meeting Notetakers Raise Legal and Privacy Risks as Enterprise Adoption Grows
Security
1
Jul 9
6
AI Meeting Notetakers Raise Legal and Privacy Risks as Enterprise Adoption Grows
Security
· 1 src · Jul 9
Discuss
6
AI Security Audit Discovers 7 Real Vulnerabilities in Cloudflare's CIRCL Cryptography Library
Security
1
Jul 7
6
AI Security Audit Discovers 7 Real Vulnerabilities in Cloudflare's CIRCL Cryptography Library
Security
· 1 src · Jul 7
Discuss
6
Reddit Deploys LLMs to Combat the AI-Generated Spam Epidemic
Security
1
Jul 6
6
Reddit Deploys LLMs to Combat the AI-Generated Spam Epidemic
Security
· 1 src · Jul 6
Discuss
3 Weeks Ago
8
Security Researcher Uses Claude to Discover Critical Vulnerability in US Music Festival Ticketing Platform
Security
1
Jul 1
8
Security Researcher Uses Claude to Discover Critical Vulnerability in US Music Festival Ticketing Platform
Top
Security
· 1 src · Jul 1
Discuss
7
Claude Code Found Embedding Steganographic Markers in System Prompts to Detect API Proxies
Security
1
Jul 1
7
Claude Code Found Embedding Steganographic Markers in System Prompts to Detect API Proxies
Top
Security
· 1 src · Jul 1
Discuss
7
Apple Breaks Patch Cadence to Counter AI-Accelerated Cyberattacks, Warns of 'Time Itself' as Attack Surface
Security
1
Jul 1
7
Apple Breaks Patch Cadence to Counter AI-Accelerated Cyberattacks, Warns of 'Time Itself' as Attack Surface
Security
· 1 src · Jul 1
Discuss
7
Researchers Demonstrate "Dream World" Attack That Bypasses AI Browser Guardrails Entirely
Security
1
Jun 30
7
Researchers Demonstrate "Dream World" Attack That Bypasses AI Browser Guardrails Entirely
Security
· 1 src · Jun 30
Discuss
7
Startup Shift Offers Free Housecleaning in Exchange for Home Video Data to Train AI Robots
Security
3
Jun 28
7
Startup Shift Offers Free Housecleaning in Exchange for Home Video Data to Train AI Robots
Security
· 3 srcs · Jun 28
Discuss
6
Amazon Bedrock Deploys Multi-Stage AI Pipeline to Detect AI-Generated Phishing Emails
Security
1
Jul 3
6
Amazon Bedrock Deploys Multi-Stage AI Pipeline to Detect AI-Generated Phishing Emails
Security
· 1 src · Jul 3
Discuss
6
LangChain Reveals WebAssembly-Based Security Architecture for AI Agent Code Execution
Security
1
Jun 30
6
LangChain Reveals WebAssembly-Based Security Architecture for AI Agent Code Execution
Security
· 1 src · Jun 30
Discuss
6
Nvidia Exec Explains How Confidential Computing Secures AI Agents on Sensitive Enterprise Data
Security
1
Jun 30
6
Nvidia Exec Explains How Confidential Computing Secures AI Agents on Sensitive Enterprise Data
Security
· 1 src · Jun 30
Discuss
Last Month
9
Anthropic Accuses Alibaba's Qwen Lab of Orchestrating Large-Scale Illicit Claude Extraction via 25,000 Fake Accounts
Updated
Security
2
Jun 25
9
Anthropic Accuses Alibaba's Qwen Lab of Orchestrating Large-Scale Illicit Claude Extraction via 25,000 Fake Accounts
Top
Security
· 2 srcs · Jun 25
Discuss
9
Anthropic's Mythos AI Cracked Classified U.S. Government Systems in Hours During Secret Testing
Security
1
Jun 24
9
Anthropic's Mythos AI Cracked Classified U.S. Government Systems in Hours During Secret Testing
Top
Security
· 1 src · Jun 24
Discuss
9
Five Eyes Intelligence Agencies Issue Rare Joint Warning: Frontier AI Cyber Threats to Governments Are "Months Away"
Updated
Security
2
Jun 29
9
Five Eyes Intelligence Agencies Issue Rare Joint Warning: Frontier AI Cyber Threats to Governments Are "Months Away"
Top
Security
· 2 srcs · Jun 29
Discuss
8
OpenAI Launches GPT-5.5-Cyber and Patch the Planet Initiative — While White House Stays Silent Despite Model Outperforming Export-Controlled Anthropic Mythos 5
Updated
Security
6
Jun 24
8
OpenAI Launches GPT-5.5-Cyber and Patch the Planet Initiative — While White House Stays Silent Despite Model Outperforming Export-Controlled Anthropic Mythos 5
Top
Security
· 6 srcs · Jun 24
Discuss
8
Meta Pauses Employee Monitoring Program After Internal Data Leak Exposes Collected Data Company-Wide
Updated
Security
7
Jun 24
8
Meta Pauses Employee Monitoring Program After Internal Data Leak Exposes Collected Data Company-Wide
Top
Security
· 7 srcs · Jun 24
Discuss
8
Critical Microsoft M365 Copilot Prompt Injection Flaw Enabled 2FA Code Theft
Security
1
Jun 16
8
Critical Microsoft M365 Copilot Prompt Injection Flaw Enabled 2FA Code Theft
Security
· 1 src · Jun 16
Discuss
8
AI Is Supercharging Cyberattacks at Unprecedented Scale as Restricted Elite Hacking Models Emerge
Security
2
Jun 16
8
AI Is Supercharging Cyberattacks at Unprecedented Scale as Restricted Elite Hacking Models Emerge
Top
Security
· 2 srcs · Jun 16
Discuss
8
Meta Licensed Pentagon-Tied Face Recognition Tech for Smart Glasses, WIRED Investigation Reveals
Security
3
Jun 15
8
Meta Licensed Pentagon-Tied Face Recognition Tech for Smart Glasses, WIRED Investigation Reveals
Top
Security
· 3 srcs · Jun 15
Discuss
8
Cornell Study: A 13-Word Reddit Comment Can Poison AI Search Outputs at Scale
Security
1
Jun 15
8
Cornell Study: A 13-Word Reddit Comment Can Poison AI Search Outputs at Scale
Top
Security
· 1 src · Jun 15
Discuss
7
LLM Coding Agents Found Exploiting Benchmarks via Web Lookup and Git History Mining
Security
1
Jun 26
7
LLM Coding Agents Found Exploiting Benchmarks via Web Lookup and Git History Mining
Top
Security
· 1 src · Jun 26
Discuss
7
Correction: NSA's Use of AI Tool Mythos Was a Controlled Red-Team Test, Not a System Breach
Security
1
Jun 24
7
Correction: NSA's Use of AI Tool Mythos Was a Controlled Red-Team Test, Not a System Breach
Security
· 1 src · Jun 24
Discuss
7
Single Fake Web Page Can Manipulate AI Shopping Recommendations, New FORGE Benchmark Reveals
Security
1
Jun 23
7
Single Fake Web Page Can Manipulate AI Shopping Recommendations, New FORGE Benchmark Reveals
Security
· 1 src · Jun 23
Discuss
7
AI-Powered Scams Surge Around 2026 FIFA World Cup, Making Fraud Harder to Detect
Security
1
Jun 22
7
AI-Powered Scams Surge Around 2026 FIFA World Cup, Making Fraud Harder to Detect
Security
· 1 src · Jun 22
Discuss
7
MosaicLeaks: AI Research Agents Expose Private Enterprise Data Through Web Query Patterns
Security
1
Jun 18
7
MosaicLeaks: AI Research Agents Expose Private Enterprise Data Through Web Query Patterns
Security
· 1 src · Jun 18
Discuss
6
AI Smart Glasses Emerge as New Exam Cheating Vector as Asia Tightens Test Security
Security
1
Jun 27
6
AI Smart Glasses Emerge as New Exam Cheating Vector as Asia Tightens Test Security
Security
· 1 src · Jun 27
Discuss
6
2,000 Hackers, 6,000 Attempts, Zero Breaches: Claude Opus 4.6 Passes Public Prompt Injection Test
Security
1
Jun 26
6
2,000 Hackers, 6,000 Attempts, Zero Breaches: Claude Opus 4.6 Passes Public Prompt Injection Test
Security
· 1 src · Jun 26
Discuss
6
Gray Swan Experts Warn Indirect Prompt Injection Breaches in AI Agents Are Likely Inevitable
Security
1
Jun 24
6
Gray Swan Experts Warn Indirect Prompt Injection Breaches in AI Agents Are Likely Inevitable
Security
· 1 src · Jun 24
Discuss
6
AI Tool Mythos Signals Shift Toward Zero-Cost Continuous Smart Contract Security Auditing
Security
1
Jun 20
6
AI Tool Mythos Signals Shift Toward Zero-Cost Continuous Smart Contract Security Auditing
Security
· 1 src · Jun 20
Discuss
6
Lithuanian Startup Launches AI-Powered Crowdsourced Smartphone Network to Detect Shahed-Type Drones
Security
1
Jun 20
6
Lithuanian Startup Launches AI-Powered Crowdsourced Smartphone Network to Detect Shahed-Type Drones
Security
· 1 src · Jun 20
Discuss
6
Benchmark Study Finds Higher LLM Reasoning Effort Doesn't Always Improve Security Vulnerability Triage
Security
1
Jun 18
6
Benchmark Study Finds Higher LLM Reasoning Effort Doesn't Always Improve Security Vulnerability Triage
Security
· 1 src · Jun 18
Discuss
Filters
Signal
Title
Category
Sources
Posted
Discuss
