Cybersecurity News | Daily Recap [24 Jul 2026]

AI & Models

• OpenAI patched a ChatGPT Agent flaw that could let attackers forge an AI insider, while industry reactions followed reports of OpenAI models being used to hack Hugging Face systems. – ChatGPT Flaw, Model Hacking
• AI-powered malware is advancing with Dolphin X, which now uses AI to rank high-value targets, while a hacker also ran the Hermes AI Agent unattended for post-exploitation at Thailand’s finance ministry. – Dolphin X, AI Targeting, Hermes Abuse
• AegisAI raised $36 million to expand AI-powered email security as defenders race to counter increasingly automated threats. – AegisAI Funding
• Agentic AI is also pressuring confidential computing progress, underscoring new trust and isolation challenges for enterprise security. – AI Risk

Ransomware & Malware

• Golden Chickens resurfaced with 4 new malware families and modular implants, signaling a broader post-compromise toolkit. – Golden Chickens
• Clop ransomware is targeting Windchill and FlexPLM in data theft attacks, continuing its extortion-focused campaign. – Clop Attack
• Attackers are abusing Notepad++ plugins to stealthily install malware, while a fake Claude app pushed via Bing ads delivered SectopRAT. – Notepad++ Abuse, Fake Claude

Espionage & Email Theft

• Russian espionage groups exploited a Zimbra zero-day to steal mail and 2FA codes, with related reporting showing a novel exploit used against Western targets. – Zimbra Zero-Day, Zimbra Theft, Western Targets

Breaches & Account Attacks

• Origin, the Australian energy giant, confirmed a data breach exposing client data after being hacked. – Origin Breach, Origin Exposes Data
• Chick-fil-A accounts were hit by a credential stuffing attack, while a man received 6 years for hacking 750 women’s Snapchat accounts. – Chick-fil-A Stuffing, Snapchat Case

Policy & Enforcement

• Europol flagged 4,340 URLs for removal in a crackdown on The Com, while Marco Rubio imposed visa restrictions on sextortionists and cyber scammers. – Europol Crackdown, Visa Restrictions

Vulnerability & Patch

• Coverage of 400 Linux kernel flaws and a broader debate over whether patching is “dead” highlight the ongoing challenge of vulnerability management. – Linux Flaws, Patching Debate

Threat Intel

• Cavalier launched new threat feeds to improve visibility into attacker infrastructure and support tracking of active campaigns. – Threat Feeds

Cybersecurity News | Daily Recap – hendryadrian.com