OpenAI Codex Security Scanned 1.2 Million Commits and Found 10,561 High-Severity Issues

2026-03-08T19:24:16Z4659bbaffe42cc21b86cea18c415055d5c5e5e1382e9b93b5813f77c0a0d34f2
AI-securityAPT28APT41 (Silver Dragon)Anthropic ClaudeAsyncRATCISA KEVClickFix campaignCoruna exploit kitDindoor backdoorDust SpecterHavoc C2LeakBase seizureLumma StealerMuddyWater (Seedworm)OpenAI CodexPackagist supply-chainTransparent TribeTycoon 2FA takedownUAT-9244VOID#GEISTXWormXenoRATiOS exploit kitphishingpost-quantum cryptography

What happened

Collection of The Hacker News reports highlighting an accelerating and diversified threat landscape: AI is being used for both defense (OpenAI Codex Security, Anthropic/Claude finding Firefox flaws) and offense (Transparent Tribe mass-producing malware). Multiple APTs and malware campaigns are active (MuddyWater with Dindoor, APT41/Silver Dragon, APT28 with BadPaw/MeowMeow, Dust Specter, China-linked UAT‑9244), plus multi-stage campaigns delivering RATs (VOID#GEIST → XWorm/AsyncRAT/Xeno RAT), Havoc C2, Lumma Stealer, and cross‑platform RATs via malicious Packagist packages. Several high‑impact

Why it matters

A reviewed impact interpretation has not been published for this record.

Evidence and limitations

Source ID
the_hacker_news
Record identifier
4659bbaffe42cc21b86cea18c415055d5c5e5e1382e9b93b5813f77c0a0d34f2
Enrichment time
2026-03-08T19:24:16Z
AI-assisted enrichment
Yes

This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.