A River of News

OpenAI agents carried out an undisclosed attack on RubyGems

40 publications, in the order they reported. The story spread over 583 hours 36 minutes. Words in highlight are ones the first headline did not use. All top stories · On NewsMarkets

Share: Bluesky · X · Facebook · LinkedIn · Email

Publications over time

23:17 UTC06:53 UTC40 publications
Publications in order of reporting
#Time (UTC)After the firstPublicationHeadline
1firstHacker NewsOpenAI agents carried out an undisclosed attack on RubyGems
2+2 hours 20 minutesThe GuardianAI agents being tested by OpenAI involved in cyber-attack on another service, say researchers by Guardian staff
3+20 hours 35 minutesEngadgetOpenAI agents hacked a software service before the Hugging Face incident
4+22 hours 24 minutesThe VergeOpenAI’s rogue AI tried to hack another company in May by Terrence O’Brien
5+66 hours 33 minutesQuartzOpenAI's rogue AI agents were attacking a code registry months before the Hugging Face breach by Cris Tolomia
6+66 hours 46 minutesThe RegisterOpenAI's malicious bot swarm attacked RubyGems
7+117 hours 13 minutesDecryptOpenAI's Rogue AI Agents Were Probing Hugging Face Two Months Before Hack by Jose Antonio Lanz
8+118 hours 43 minutesAxiosOpenAI discloses six new safety incidents by Ina Fried
9+148 hours 14 minutesFinancial TimesOpenAI breached by researchers using Anthropic models
10+158 hours 13 minutesArs TechnicaResearchers used Claude to hack OpenAI by Cristina Criddle, Financial Times
11+158 hours 27 minutesTom's HardwareHackers breach OpenAI using Claude tools, gaining access to employee accounts and the company's internal codebase — attackers initiated a 'harmless' pull request as proof of the hack by Etiido Uko
12+167 hours 28 minutesBloombergGoogle Joins OpenAI, Anthropic, Meta in Disclosing AI Hacks by Julia Love, Davey Alba
13+168 hours 5 minutesThe InformationGoogle’s Gemini Model Hacks Companies During Test by Nick Wingfield
14+170 hours 20 minutesAl JazeeraGoogle’s Gemini AI hacks 3 companies in security test, then stops
15+172 hours 18 minutesDeutsche WelleGoogle's Gemini AI hacked 3 companies during testing
16+172 hours 45 minutesSouth China Morning PostGoogle says Gemini AI model breached real systems in security test
17+176 hours 24 minutesSky NewsGoogle's Gemini AI hacks three other companies during security test
18+186 hours 12 minutesTechCrunchGoogle’s Gemini is the latest AI model to hack other companies by Anthony Ha
19+213 hours 3 minutesMarkTechPostYou too Google! Google Confirms Gemini Breached 3 Companies in AI Security Tests by Asif Razzaq
20+273 hours 42 minutesMIT Technology ReviewThe AI Hype Index: AI loves cheating by Michelle Kim
21+286 hours 53 minutesFrance 24OpenAI AI agent breached Australian government website, PM says by FRANCE 24
22+288 hours 42 minutesNatureAI agent hacks government website for first time: why this breach matters by Mohana Basu
23+289 hours 54 minutesThe Japan TimesOpenAI agent hacked Australian government website, Albanese says same headline as Bloomberg
24+290 hours 21 minutesBusiness InsiderOops. OpenAI's bot hacked an Australian government website. by Aditi Bharade
25+290 hours 51 minutesFortuneOpenAI’s agent hacked Australia’s Medicare website—the latest rogue AI incident that the company didn’t know about for months by Emily Forlini
26+291 hours 40 minutesTimes of India'Obviously unacceptable': Australia accuses OpenAI of breaching government website by ROHITASHWA RANJAN
27+293 hours 56 minutesNPROpenAI's breach of Australian health department website prompts rebuke by The Associated Press
28+297 hours 9 minutesNew ScientistOpenAI agent hacked an Australian government healthcare website by James Woodford
29+297 hours 11 minutesCointelegraphAustralia says OpenAI agent hacked government site before Altman warning by Cointelegraph by Helen Partz
30+298 hours 4 minutesCBS NewsAustralia says rogue OpenAI model hacked into its healthcare system
31+299 hours 29 minutesWIREDAn OpenAI Agent Hacked Australia’s Health Service. Their Government Found Out Months Later. by Isabella Ward
32+302 hours 16 minutesMedPage TodayAI Hacks Health Network; White House Drops NIH Plan; RFK Jr. Makes Millions off MAHA
33+303 hours 27 minutesScientific AmericanOpenAI’s agent hacking Australia is a warning for governments everywhere
34+303 hours 36 minutesThe HillAlbanese says OpenAI agent hacked Australia health department site by Ryan Mancini
35+305 hours 17 minutesCNETAustralia Says an OpenAI Agent Hacked Into a Goverment Health Site by Dashia Milden
36+378 hours 36 minutesMother JonesRest Assured: AI Companies Say They’re Investigating Tens of Thousands of Rogue Bot Incidents by Alex Nguyen
37+409 hours 42 minutesOpenAIHow we will do better for Australia
38+443 hours 26 minutesCoinDeskOpenAI, Google and Meta pledge outside AI audits under voluntary White House deal by Shaurya Malwa
39+579 hours 14 minutesThe New York TimesOpenAI’s models have also meddled with U.S. government websites after going rogue. by Kate Conger, Ana Swanson, Cecilia Kang
40+583 hours 36 minutesBBCOpenAI admits response to Australian government hacks 'not good enough'

Who covered it, by audience

19 of 40 publications on this story have an audience score.

Publications by the political lean of their audience
Audience leans Democratic (10)Mixed audience (8)Audience leans Republican (1)

Audience lean comes from Robertson, R. E., Jiang, S., Joseph, K., Friedland, L., Lazer, D., and Wilson, C. (2018), Auditing Partisan Audience Bias within Google Search, Proceedings of the ACM on Human-Computer Interaction 2(CSCW), article 148, used with the authors' permission. The score shows which party's registered voters, on Twitter, shared an outlet's links more. It is not a rating of the outlet's own slant, and it dates from 2018. See how we use it.