Wall Street Journal · Robert McMillan ·
Security researchers in OpenAI's bug bounty program hacked OpenAI in July and accessed its "monorepo" on GitHub, using Opus 4.8 for cybersecurity and Opus 5
A bug-hunting independent security research team was able to access OpenAI's internal code system, exposing growing risks in automated cyber threats
Lead Source
How this story grew
Coverage · 0
Discussion · 0
Sep 18Sep 19
Coverage
More
Hacktron AI: Hacktron AI
New York Times: New York Times
The Information: The Information
CNBC: CNBC
SiliconANGLE: SiliconANGLE
The Guardian: The Guardian
TechCrunch: TechCrunch
TechRadar: TechRadar
The Coin Republic: The Coin Republic
CyberScoop: CyberScoop
NBC News: NBC News
PCMag: PCMag
Proactive: Proactive
International Business Times: International Business Times
NewsMax.com: NewsMax.com
Gizmodo: Gizmodo
The Decoder: The Decoder
The Indian Express: The Indian Express
The Verge: The Verge
PYMNTS: PYMNTS
Quartz: Quartz
AI Policy Daily: AI Policy Daily
The Neuron: The Neuron
Metacurity: Metacurity
Financial Times: Financial Times
Discussion
@s1r1u5_ @andrewcurran_ @r0bre @s1r1u5_ @cramforce @iminurputer @isaacking314 @jeremiahdillon @weezerosint @mikeisaac @linchzhang @florian_jue @drelidavid @leahmcelrath @artemr @zackkorman @nickadobos @mikeisaac @hlntnr @jeffladish @garymarcus @yuchenj_uw @enginoid @reedalbergotti @kpolley @tracewoodgrains @symbolsrsymbols @aisafetymemes @notdeghost @sydneyvonarx @justanotherlaw @s1r1u5_ @s1r1u5_ @s1r1u5_ @s1r1u5_ @joshua_saxe @justanotherlaw
Related stories
Wall Street Journal: Researchers: OpenAI agents attacked Ruby package manager RubyGems in May; OpenAI says its agents used RubyGems to access the internet to do "benign tasks"