OpenAI Reveals GPT-5.6 Sol Cybersecurity Model, Restricts Early Access
2026-06-29T20:52:14Z•6b1288f14bf1fd6a100569595aa59feca4869d99cb4ed6cbc2189aa79dcae97f
AI modelsApple BootROMChina-linkedCisco Catalyst SD-WANEDR bypassFortiBleed guidance」、「supply chain attack」「Mastra」「North Korea」「FortinetGPT-5.6Jaguar Land RoverMDM bypassMillenium RATNAIC breachOpenAIOracle PeopleSoftRussiaTelegramTinyRCTcritical infrastructuredestructive attackmacOS vulnerabilitymalwarenation-statepre-disclosure exploitationunpatchable exploitzero-day
What happened
A roundup of cybersecurity news: OpenAI previewed its GPT-5.6 Sol model under restricted access at the US government's request. Multiple large-scale malware and espionage incidents were reported — a Telegram-based Millenium RAT campaign infected ~62k devices worldwide, China-linked TinyRCT backdoors targeted Southeast Asian critical infrastructure, and macOS/iOS flaws (including an unpatchable BootROM exploit for A12/A13) and a macOS XPC bug that can disable EDR/MDM put endpoint security at risk. Several high-impact breaches and vulnerabilities were disclosed or exploited: an attacker used a 0
Why it matters
A reviewed impact interpretation has not been published for this record.
Evidence and limitations
- Source ID
- infosecurity_magazine
- Record identifier
- 6b1288f14bf1fd6a100569595aa59feca4869d99cb4ed6cbc2189aa79dcae97f
- Enrichment time
- 2026-06-29T20:52:14Z
- 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.