---
title: "Researchers Hack OpenAI Using Anthropic AI Model"
publication: "MediaBias News"
url: "https://mediabias.news/business/researchers-hack-openai-using-anthropic-ai-model"
api: "https://mediabias.news/api/v1/stories/researchers-hack-openai-using-anthropic-ai-model"
markdown: "https://mediabias.news/business/researchers-hack-openai-using-anthropic-ai-model.md"
audio: "https://mediabias.news/api/audio/researchers-hack-openai-using-anthropic-ai-model"
category: "Business & Economy"
published: "2026-09-18T17:05:12.852Z"
source_reported: "2026-09-18T12:34:55.000Z"
updated: "2026-09-18T17:05:12.852Z"
trust_score: 50
critic_score: 65
hype_score: 45
assessment_type: "coverage-and-source-reporting-analysis"
fact_check_status: "not-performed"
coverage_measured: "2026-09-18T16:44:51.360130+00:00"
---

# Researchers Hack OpenAI Using Anthropic AI Model

*Cybersecurity team exploited a forum vulnerability to access internal code, earning a bug bounty.*

**Scores for the source reporting** (0-100, assessing the original journalism this article was written from, not this write-up): 

- Trust 50 of 100, higher is better. Central claim attributed to named body (Hacktron AI) and corroborated by two independent outlets.
- Craft 65 of 100, higher is better. Reports quote OpenAI and name the researchers, and specify the timeline and method of the breach.
- Hype 45 of 100, lower is better. Headlines emphasize the 'hack' and the use of a competitor's AI, but the body details a bug bounty program.

Scored by MediaBias News; method at https://mediabias.news/methodology.

**The short version**

- Cybersecurity researchers accessed OpenAI's internal systems using Anthropic's Claude AI model.
- The hack exploited a vulnerability in OpenAI's community forum related to image processing.
- Researchers gained access to an OpenAI employee's ChatGPT account, linked to the company's code repository.
- OpenAI awarded the researchers $6,500 for discovering and reporting the vulnerabilities.

Cybersecurity researchers from Hacktron AI successfully accessed internal systems at OpenAI by leveraging Anthropic's Claude AI model. The incident, which occurred in July, involved exploiting a vulnerability in OpenAI's community forum that was related to image processing. This allowed the researchers to gain access to an OpenAI employee's ChatGPT account, which was connected to the company's code repository. The details of the breach were later revealed by The Wall Street Journal. OpenAI acknowledged the findings and rewarded the Hacktron AI team with $6,500 for identifying and reporting the security flaws.

## Disagreement on timeline

Tech Radar reported that the entire timeline from vulnerability discovery to repository access took less than 72 hours. Eldiario.es stated the attack began on 23 July and that the researchers revealed the details exclusively to The Wall Street Journal now, without specifying a total duration for the exploit.

## What the coverage left out

None of the left-rated reports below mention the specific vulnerability related to the libheif exploit or the ImageMagick conversion process that was detailed in the centre-rated Tech Radar report. None of the right-rated reports below mention the specific vulnerability related to the libheif exploit or the ImageMagick conversion process that was detailed in the centre-rated Tech Radar report.

## Who covered it

Shares of the 45 covering outlets with a published leaning rating:

- Left: 18% (8)
- Centre: 40% (18)
- Right: 42% (19)

79 of 124 covering outlets carried no usable leaning rating and were excluded from those percentages. Coverage measured 2026-09-18T16:44:51.360130+00:00.

**Coverage watch:** developing. Checked 2 times; 0 checks found a material change. Most recently checked 2026-09-18T17:30:08.872Z.

## How the sides framed it

### Left

- Reports from the left led with the fact that researchers used Anthropic's Claude AI to hack into OpenAI's systems. Both Eldiario.es and Business Insider highlighted that the Hacktron AI team was rewarded $6,500 for their discovery. Eldiario.es also noted that the hack occurred in July and was revealed to The Wall Street Journal, and that the researchers used Claude Opus 5 after initial attempts with Opus 4.8 failed. Business Insider described Hacktron as a 'tiny cybersecurity startup'.

### Centre

- Centre-rated reports focused on the breach itself and the speed at which it occurred. Tech Radar stated the hack took less than 72 hours from discovery to repository access, and that the researchers used Claude Opus 5 after initial struggles with Opus 4.8. Quartz noted the three-person Hacktron AI team chained two vulnerabilities to reach OpenAI's private GitHub environment in under 72 hours. Fortune reported that three individuals using Claude hacked into OpenAI and accessed its source code for a $6,500 reward. The Wall Street Journal described the incident as exposing 'growing risks in automated cyber threats'.

### Right

- Right-rated reports highlighted that researchers used Claude, a competitor's AI, to breach OpenAI. Zero Hedge stated that three security researchers used Anthropic's Claude to breach OpenAI employee accounts and access private company software. Protothema reported that the hack raised 'new concerns about protecting AI models'. India Today noted that Indian-origin researchers hacked OpenAI using Claude and were paid approximately Rs 6.2 lakh (equivalent to $6,500). Latestly mentioned that researchers used Claude to exploit vulnerabilities in OpenAI's community forum, gaining access to an employee's ChatGPT and Codex environment and linked GitHub repositories, demonstrating the breach with a harmless pull request. Voz.us quoted Hacktron AI's CTO, Mohan Pedhapati, saying, 'I don't think we're as strong as Chinese threat actors. We're just three guys with subscriptions to Claude and Codex.'

## Factuality profile of the covering outlets

These are published factuality ratings of the outlets, not a verdict on whether this story or its claims are true.

- veryLow: 1
- low: 1
- mixed: 7
- high: 25
- unknown: 79
- veryHigh: 11

## Verification scope

This page compares coverage and scores the source reporting. It is not a ClaimReview verdict on whether the underlying event or claim is true.

## Original reporting this was written from

- [La Opinion](https://laopinion.com/2026/09/18/tres-investigadores-lograron-acceder-a-openai-en-menos-de-72-horas-con-ayuda-de-claude) — Three Researchers Managed to Access OpenAI in Less than 72 Hours with the Help of Claude
- [La Nación, Grupo Nación](https://nacion.com/el-mundo/usaron-una-inteligencia-artificial-de-la/XW7OQKTSLNHLDKJNUP5S2EPLA4/story) — They Used Artificial Competition Intelligence and Managed to Breach ChatGPT in Three Hours.
- [Times of India](https://timesofindia.indiatimes.com/world/us/meet-harsh-jaiswal-mohan-pedhapati-and-rahul-maini-3-indian-origin-researchers-who-used-claude-to-breach-openai-systems-won-6500-bounty/articleshow/134338838.cms) — Meet Harsh Jaiswal, Mohan Pedhapati and Rahul Maini: 3 Indian-origin researchers who used Claude to breach OpenAI systems, won $6,500 bounty
- [elDiarioAR](https://eldiarioar.com/tecnologia/equipo-investigadores-logra-acceder-algoritmos-secretos-openai-usando-ia-anthropic_1_13521648.html) — A Team of Researchers Gain Access to OpenAI Secret Algorithms Using Anthropic AI
- [insider.gr](https://insider.gr/epiheiriseis/425341/ereynites-hakaran-ithika-tin-openai-me-ti-boitheia-toy-claude-tis-anthropic) — Researchers Ethically Hacked OpenAI with the Help of Anthropic's Claude
- [The Verge](https://theverge.com/ai-artificial-intelligence/997444/openai-hack-claude-heif-heist) — Security researchers used Claude to help them hack into OpenAI
- [cyprustimes.com](https://cyprustimes.com/tech-news/chakaran-techniti-noimosyni-me-ergaleio-technitis-noimosynis) — They Hacked Artificial Intelligence with... an Artificial Intelligence Tool!
- [rthess.gr](https://rthess.gr/epistimi-ygeia-technologia/ai-enantion-ai-montelo-tis-anthropic-apektise-prosvasi-se-logariasmous-tis-openai) — AI "Against" AI / Anthropic Model Gained Access to OpenAI Accounts
- [TVN](https://tvn-2.com/mundo/insolito-claude-anthropic-hackear-openai_1_2261597.html) — Unusual: They Use Anthropic's "Claude" to Hack OpenAI
- [PressNewsAgency](https://pressnewsagency.org/researchers-used-claude-to-breach-openais-internal-systems) — Researchers used Claude to breach OpenAI’s internal systems
- [reader.gr](https://reader.gr/tehnologia/ereynites-hakaran-openai-hrisimopoiontas-ergaleio-tehnitis-noimosynis-tis-anthropic/662051) — Researchers Hacked OpenAI Using Anthropic's Artificial Intelligence Tool
- [IndexHR](https://index.hr/vijesti/clanak/testirali-sigurnost-openaija-i-uspjeli-uci-u-sustav-pomogao-im-je-claude/2836117.aspx) — Tested OpenAI's Security and Managed to Break Into the System. Claude Helped Them
- [Hot News](https://hotnews.ro/cercetatorii-au-folosit-claude-pentru-a-sparge-sistemele-openai-profitand-de-doua-vulnerabilitati-critice-2352738) — Researchers Used Claude to Break openAI Systems, Taking Advantage of Two Critical Vulnerabilities - HotNews.ro
- [Newsit.gr](https://newsit.gr/kosmos/ereynites-xakaran-tin-openai-me-ti-voitheia-tou-claude/4777809) — Researchers "Hacked" OpenAI with the Help of Claude
- [nafeza2world.com](https://nafeza2world.com/Akhbar-Al-Alm/2786734) — A Window of Researchers Hacking Into Chat-GPT Using Claude.
- [charentelibre.fr](https://charentelibre.fr/sciences-et-technologie/intelligence-artificielle/des-chercheurs-ont-pirate-openai-en-utilisant-une-ia-d-anthropic-30686802.php) — Researchers Hacked OpenAI Using an AI of Anthropic
- [Lietuvos Radijas ir Televizija](https://lrt.lt/naujienos/mokslas-ir-it/11/3056404/tyrejai-pasitelke-claude-kad-isilauztu-i-bendroves-openai-vidines-sistemas) — Researchers Used Claude to Hack Into OpenAI's Internal Systems
- [elfinancierocr.com](https://elfinancierocr.com/tecnologia/investigadores-usan-ia-de-anthropic-para-vulnerar/LEKP7FMKOZGSTCIWB237RE33NQ/story) — Researchers Use Anthropic AI to Breach OpenAI Employee Accounts
- [Liberation](https://liberation.fr/sciences/technologie-innovation/openai-pirate-avec-succes-par-le-chatbot-danthropic-lors-dun-exercice-de-securite-20260918_TH3ZLVJCAJHFVFRBPJ7NJ6FCKQ) — OpenAI Successfully Hacked by Anthropic's Chatbot During a Safety Exercise
- [sigmalive.com](https://sigmalive.com/news/scitech/1329991/erefnites-kivernoasfaleias-khakaran-tin-openai-me-erghalio-tis-anthropic) — Cybersecurity Researchers Hacked OpenAI with Anthropic Tool
- [Sapo](https://pplware.sapo.pt/inteligencia-artificial/investigadores-conseguiram-entrar-nos-sistemas-da-openai-com-ajuda-do-claude-da-anthropic) — Researchers Were Able to Access OpenAI Systems with the Help of Claude From Anthropic
- [news-pravda.com](https://news-pravda.com/world/2026/09/18/2601918.html) — OpenAI has been hacked. The irony of the year in the tech world: ChatGPT developers were hacked with the help of their main competitor, the Claude neural network from Anthropic
- [Clarin](https://clarin.com/estados-unidos/investigadores-afirman-haber-hackeado-openai-usando-la-ia-claude-de-anthropic_0_5GmZeldZ3i.html) — Researchers Claim to Have Hacked OpenAI Using Anthropic's AI Claude
- [The CSR Journal](https://thecsrjournal.in/indian-origin-researchers-use-claude-ai-to-hack-openai-exposing-security-flaws) — Indian-Origin Researchers Use Claude AI to Hack OpenAI, Exposing Security Flaws
- [Sud Ouest](https://sudouest.fr/sciences-et-technologie/intelligence-artificielle/chatgpt-des-chercheurs-ont-pirate-openai-en-utilisant-une-ia-d-anthropic-30685130.php) — OpenAI: Researchers Hacked Employees' ChatGPT Accounts Using an IA fromAnthropic
- [The Straits Times](https://straitstimes.com/world/united-states/researchers-used-claude-to-breach-openais-internal-systems) — Researchers used Claude to breach OpenAI’s internal systems
- [RTL Info](https://rtl.be/actu/monde/international/des-chercheurs-mettent-openai-lepreuve-ils-decouvrent-une-porte-dentree-vers-les/2026-09-18/article/800646) — Researchers Test OpenAI: They Discover a Gateway to Employee Accounts
- [StartupTalky](https://startuptalky.com/news/anthropics-claude-ai-allegedly-used-by-hackers) — Claude Used to Hack OpenAI Employee Account: What Happened
- [Daily Sabah](https://dailysabah.com/business/economy/anthropics-claude-used-to-breach-openais-internal-systems) — Anthropic's Claude used to breach OpenAI's internal systems
- [bienpublic.com](https://bienpublic.com/science-et-technologie/2026/09/18/une-faille-trouvee-en-trois-heures-openai-pirate-grace-a-l-ia-d-anthropic) — Internet. A Flaw Found "in Three Hours": OpenAI Hacked Through Anthropic AI
- [estrepublicain.fr](https://estrepublicain.fr/science-et-technologie/2026/09/18/une-faille-trouvee-en-trois-heures-openai-pirate-grace-a-l-ia-d-anthropic) — Internet. A Flaw Found "in Three Hours": OpenAI Hacked Through Anthropic AI
- [lalsace.fr](https://lalsace.fr/science-et-technologie/2026/09/18/une-faille-trouvee-en-trois-heures-openai-pirate-grace-a-l-ia-d-anthropic) — Science. A Flaw Found "in Three Hours": OpenAI Hacked Through Anthropic AI
- [ledauphine.com](https://ledauphine.com/science-et-technologie/2026/09/18/une-faille-trouvee-en-trois-heures-openai-pirate-grace-a-l-ia-d-anthropic) — Science. A Flaw Found "in Three Hours": OpenAI Hacked Through Anthropic AI
- [lejsl.com](https://lejsl.com/science-et-technologie/2026/09/18/une-faille-trouvee-en-trois-heures-openai-pirate-grace-a-l-ia-d-anthropic) — Science. A Flaw Found "in Three Hours": OpenAI Hacked Through Anthropic AI
- [leprogres.fr](https://leprogres.fr/science-et-technologie/2026/09/18/une-faille-trouvee-en-trois-heures-openai-pirate-grace-a-l-ia-d-anthropic) — Technology. A Fault Found "in Three Hours": OpenAI Hacked Through Anthropic AI
- [republicain-lorrain.fr](https://republicain-lorrain.fr/science-et-technologie/2026/09/18/une-faille-trouvee-en-trois-heures-openai-pirate-grace-a-l-ia-d-anthropic) — Internet. A Flaw Found "in Three Hours": OpenAI Hacked Through Anthropic AI
- [vosgesmatin.fr](https://vosgesmatin.fr/science-et-technologie/2026/09/18/une-faille-trouvee-en-trois-heures-openai-pirate-grace-a-l-ia-d-anthropic) — Science. A Flaw Found "in Three Hours": OpenAI Hacked Through Anthropic AI
- [Punto Informatico](https://punto-informatico.it/ricercatori-hanno-usato-claude-colpire-openai) — Researchers Used Claude to Hit OpenAI
- [Le Figaro](https://lefigaro.fr/secteur/high-tech/des-chercheurs-ont-pirate-openai-en-utilisant-une-ia-d-anthropic-20260918) — Researchers Hacked OpenAI Using an Anthropic AI
- [telex.hu](https://telex.hu/techtud/2026/09/18/openai-anthropic-claude-chatbot-hekkeles) — OpenAI Systems Were Hacked Using Anthropic's Claude Chatbot and Its Own AI Model
- [transtelex](https://transtelex.ro/techtud/2026/09/18/openai-anthropic-claude-chatbot-hekkeles) — OpenAI Systems Were Hacked Using Anthropic's Claude Chatbot and Its Own AI Model
- [01net](https://01net.com/actualites/openai-a-ete-pirate-grace-a-lia-de-son-grand-rival.html) — OpenAI Has Been Hacked Thanks to the AI of Its Great Rival
- [TechCrunch](https://techcrunch.com/2026/09/18/researchers-used-anthropics-claude-to-hack-into-openai) — Researchers used Anthropic's Claude to hack into OpenAI
- [Insider Paper](https://insiderpaper.com/researchers-used-claude-to-breach-openais-internal-systems) — Researchers used Claude to breach OpenAI's internal systems
- [Tom's Hardware](https://tomshardware.com/tech-industry/cyber-security/hackers-breach-openai-using-claude-tools-gaining-access-to-employee-accounts-and-the-companys-internal-codebase-initiating-a-harmless-pull-request-as-proof-of-the-hack) — Hackers 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
- [techloy.com](https://techloy.com/openai-hacked-using-claude) — OpenAI Hacked Using Claude: How It Happened
- [Rzeczpospolita](https://cyfrowa.rp.pl/ai/art45160251-claude-wlamal-sie-do-chatgpt-hakerzy-buszowali-w-najbardziej-znanej-firmie-ai) — Claude Hacked ChatGPT. Hackers Raided the World's Most Famous AI Company.
- [Managers](https://managers.tn/2026/09/18/cybersecurite-claude-plus-performant-que-chatgpt) — Cybersecurity: Claude Better than ChatGPT?
- [La Libre](https://lalibre.be/economie/entreprises-startup/2026/09/18/des-chercheurs-ont-pirate-openai-en-utilisant-une-ia-danthropic-BNOFOLFXHBEDHGKOJUJF63LARI) — Researchers Hacked OpenAI Using an Anthropic AI
- [Ars Technica](https://arstechnica.com/ai/2026/09/researchers-used-claude-to-hack-openai) — Researchers used Claude to hack OpenAI
- [almasryalyoum.com](https://almasryalyoum.com/news/details/4362443) — With the Help of the Claude... Researchers Hacking Into OpenAI and Reaching the Source Code for ChatGPT Today.
- [Crypto Briefing](https://cryptobriefing.com/researchers-exploit-openai-vulnerabilities-using-anthropics-ai-wsj) — Researchers exploit OpenAI vulnerabilities using Anthropic's AI: WSJ
- [RPP](https://rpp.pe/mundo/actualidad/un-equipo-de-investigadores-jaqueo-a-openai-con-claude-la-ia-de-anthropic-segun-wall-street-journal-noticia-1708158) — A Team of Researchers Checked OpenAI with Claude, Anthropic AI, According to Wall Street Journal
- [Cryptopolitan](https://cryptopolitan.com/researchers-breach-openai-anthropic-claude) — Researchers earn $6,500 for OpenAI breach using Anthropic's Claude
- [Cryptoast](https://cryptoast.fr/claude-anthropic-utilise-pour-pirater-code-source-openai) — Hackers Use Claude and GPT-5.6 Sol to Hack OpenAI
- [securityweek.com](https://securityweek.com/ai-built-exploit-and-sign-in-flaw-opened-path-to-internal-openai-code) — AI-Built Exploit and Sign-In Flaw Opened Path to Internal OpenAI Code
- [O Cafezinho](https://ocafezinho.com/2026/09/18/tres-pesquisadores-invadiram-a-openai-com-ajuda-de-uma-ia-rival-e-chegaram-ao-codigo-interno-em-72-horas) — Three Researchers Invaded OpenAI with the Help of a Rival AI and Arrived at the Internal Code in 72 Hours
- [Eldiario.es](https://eldiario.es/tecnologia/equipo-investigadores-logra-acceder-algoritmos-secretos-openai-usando-ia-anthropic_1_13520376.html) — A Team of Researchers Gain Access to OpenAI Secret Algorithms Using Anthropic AI
- [frognews.bg](https://frognews.bg/novini/probiha-openai-instrumenti-nai-golemiia-konkurent-anthropic.html) — OpenAI Was Hacked with Tools From Its Biggest Competitor Anthropic
- [Flying Eze](https://flyingeze.com/claude-openai-hacktron-bug-bounty-security-breach) — Claude Helped Researchers Breach OpenAI in Security Test
- [The Express Tribune](https://tribune.com.pk/story/2630048/anthropics-claude-helped-cybersecurity-researchers-breach-openai-report) — Anthropic's Claude helped cybersecurity researchers breach OpenAI: report
- [Quartz](https://qz.com/hacktron-claude-openai-hack-internal-repository-091826) — Hackers used Claude to break into OpenAI's internal code repo
- [indyturk.com](https://indyturk.com/node/782665) — A Security Vulnerability Has Been Discovered in OpenAI's System.
- [computersweden.se](https://computersweden.se/article/4223792/forskare-hackade-open-ai-med-hjalp-av-claude.html) — Researchers Hacked Open AI with the Help of Claude
- [NEWS9 LIVE](https://news9live.com/technology/artificial-intelligence/heif-heist-claude-ai-photo-openai-hack-researchers-3008773) — HEIF Heist: How three researchers used Claude AI and a photo to hack OpenAI’s internal systems
- [Zero Hedge](https://zerohedge.com/ai/researchers-used-claude-hack-openai-employee-accounts) — Researchers Used Claude To Hack OpenAI Employee Accounts
- [TV News Check](https://tvnewscheck.com/ai/article/hackers-used-anthropics-claude-to-break-into-openai) — Hackers Used Anthropic’s Claude To Break Into OpenAI
- [The Tech Portal](https://thetechportal.com/2026/09/18/openai-hacked-using-claude-hacktron-ai-indian-security-researchers) — Three Indian researchers used Claude to hack into OpenAI in under 72 hours
- [Folha de S.Paulo](https://redir.folha.com.br/redir/online/emcimadahora/rss091/*https://c-level.folha.uol.com.br/ia/2026/09/hackers-do-bem-invadem-openai-usando-programa-da-rival-anthropic-diz-dona-do-chatgpt.shtml) — 'Hackers of Good' Invade OpenAI Using an Anthropic Rival Program, Says the Owner of ChatGPT - 18/09/2026
- [LiFO](https://lifo.gr/now/tech-science/ereynites-parabiasan-ta-systimata-tis-openai-hrisimopoiontas-logismiko-tis) — Researchers Hacked OpenAI Systems Using Anthropic Software
- [DiarioBitcoin](https://diariobitcoin.com/blockchain/claude-opus-5-ayudo-a-hacktron-ai-a-infiltrarse-en-openai-en-menos-de-72-horas) — Claude Opus 5 Helped Hacktron AI Infiltrate OpenAI in Less than 72 Hours
- [Crypto Breaking News](https://cryptobreaking.com/hacktron-says-it-hacked-openai) — Hacktron Says It Hacked Openai in Under 72 Hours
- [Protothema](https://protothema.gr/world/article/1880810/ereunites-kuvernoasfaleias-hakaran-tin-openai-me-ergaleio-tis-anthropic-nees-anisuhies-gia-tin-prostasia-ton-modelon-ai) — Cybersecurity Researchers Hack OpenAI with Anthropic Tool: New Concerns About Protecting AI Models
- [La Razón](https://larazon.es/tecnologia/grupo-hackers-uso-claude-entrar-openai-eso-bueno_202609186aad1898c2f1e923640fc8c6.html) — A Bunch of Hackers Used Claude to Get Into OpenAI, and that's Good.
- [Tech Radar](https://techradar.com/pro/security/white-hat-hackers-just-breached-openai-using-anthropics-claude-in-less-than-72-hours-and-it-is-a-case-study-in-just-how-fast-ai-is-advancing) — White hat hackers just breached OpenAI using Anthropic's Claude in less than 72 hours — and it is a case study in just how fast AI is advancing
- [abc](https://abc.es/tecnologia/logran-hackear-openai-horas-utilizando-ia-anthropic-20260918124846-nt.html) — They Achieve 'Hacking' OpenAI in 72 Hours Using the Anthropic AI
- [Merca2.es](https://merca2.es/2026/09/18/hackers-claude-anthropic-hackeo-openai-2457451) — Hackers Use Anthropic Claude to Hack OpenAI: Alert for Malicious Use of AI
- [India Today](https://indiatoday.in/technology/news/story/indian-origin-researchers-hack-openai-using-claude-get-paid-rs-6-lakh-2997681-2026-09-18) — Indian-origin researchers hack OpenAI using Claude, get paid Rs 6 lakh
- [Bitcoin.com](https://news.bitcoin.com/security/openai-hacked-white-hat-researchers-anthropic-claude-opus-5) — White Hats Used Anthropic's Claude to Break Into OpenAI in 72 Hours
- [δημοκρατία](https://dimokratia.gr/kosmos/746552/o-polemos-ton-ai-ereynites-chakaroyn-tin-openai-me-logismiko-tis-anthropic) — AI War: Researchers Hack OpenAI with Anthropic Software
- [Business Insider (Poland)](https://businessinsider.com.pl/technologie/nowe-technologie/badacze-wlamali-sie-do-openai-z-pomoca-anthropic-wykryto-powazna-luke/e8pkde5) — Another AI Blunder: OpenAI Was Attacked by Researchers Using Anthropic Models.
- [Reformatorisch Dagblad](https://rd.nl/a/ethische-hackers-breken-binnen-72-uur-in-bij-chat-gpt-maker-open-ai) — Ethical Hackers Break Into ChatGPT Maker OpenAI Within 72 Hours
- [Numerama](https://numerama.com/cyberguerre/2335357-comment-claude-opus-5-a-aide-des-chercheurs-a-acceder-aux-depots-internes-dopenai.html) — How Claude Opus 5 Helped Researchers Access OpenAI's Internal Repositories
- [TechJuice](https://techjuice.pk/hacktron-researchers-used-claude-breach-openai-employee-github-bounty) — Researchers Hacked OpenAI Using Anthropic's Claude AI Model
- [Haberler](https://haberler.com/guncel/openai-in-kullandigi-sistemdeki-guvenlik-acigi-20249690-haberi) — The Security Vulnerability in the System Used by OpenAI Was Discovered with Anthropic's Claude Software.
- [Fortune](https://fortune.com/2026/09/18/open-ai-hacked-anthropic-claude-source-code-6500-reward) — Three guys using Anthropic’s Claude hacked into OpenAI and accessed its source code for $6,500 reward
- [Nieuws.nl](https://nieuws.nl/economie/ethische-hackers-breken-binnen-72-uur-in-bij-chatgpt-maker-openai) — Ethical Hackers Break Into ChatGPT Maker OpenAI Within 72 Hours
- [welingelichtekringen.nl](https://welingelichtekringen.nl/economie/ethische-hackers-breken-binnen-72-uur-in-bij-chatgpt-maker-openai) — Ethical Hackers Break Into ChatGPT Maker OpenAI Within 72 Hours
- [Latestly](https://latestly.com/technology/is-it-easy-to-hack-openai-know-how-researchers-did-it-using-anthropics-claude-7609904.html) — Is It Easy To Hack OpenAI? Know How Researchers Did It Using Anthropic’s Claude | 📲 LatestLY
- [voz.us](https://voz.us/es/tecnologia-ciencia/260918/39210/crisis-ciberseguridad-ia-da-tregua-usaron-claude-anthropic-hackear-openai-semanas-despues-propio-ataque-hugging-face.html) — The Cyber-Security Crisis in the AI Is No Truce: Three Investigators Used Claude From Anthropic to Hack OpenAI Weeks After Their Own Attack on Hugging Face
- [Business Insider (Spain)](https://businessinsider.es/tecnologia/esta-pequena-empresa-hackeo-openai-usando-claude-gano-6-500-dolares_7038431_0.html) — This Small Company Hacked OpenAI Using Claude and Won $6,500
- [Expansión](https://expansion.com/economia/financial-times/2026/09/18/6aad0560468aeb5c408b4575.html) — Researchers Manage to Violate OpenAI Using Anthropic Models
- [it-daily.net](https://it-daily.net/shortnews/hacker-claude-openai) — Hackers Used Claude for Breaking Into OpenAI
- [Aroged](https://aroged.com/2026/09/18/using-anthropic-ai-models-to-attack-openai-systems) — Using Anthropic AI models to attack OpenAI systems
- [The Information](https://theinformation.com/briefings/bug-hunters-used-claude-hack-openai) — Bug Hunters Used Claude to Hack OpenAI
- [Wall Street Journal](https://wsj.com/tech/ai/hackers-used-anthropics-claude-to-break-into-openai-b40ba883) — Hackers Used Anthropic’s Claude to Break Into OpenAI
- [The Cryptonomist](https://cryptonomist.ch/2026/09/18/hacking-con-intelligenza-artificiale) — Hacking with Artificial Intelligence: as Hacktron Ai Hacktron Hacked Openai
- [Blockonomi](https://blockonomi.com/anthropics-claude-ai-exploited-to-infiltrate-openai-systems-within-72-hours) — Anthropic's Claude AI Exploited to Infiltrate OpenAI Systems Within 72 Hours
- [Aleph News](https://alephnews.ro/tehnologie/cercetatorii-in-securitate-cibernetica-au-reusit-sa-acceseze-sistemele-interne-ale-openai-folosind-claude) — Researchers Have Managed to Access OpenAI Systems Using Claude
- [Handelsblatt](https://handelsblatt.com/technik/ki/cybersicherheit-hacker-dringen-offenbar-mit-anthropic-ki-in-openai-ein/100255565.html) — Cybersecurity: Hackers Appear to Penetrate OpenAI with Anthropic-KI
- [Coin Academy](https://coinacademy.fr/actu/openai-hack-pirate-claude-ia-anthropic) — OpenAI Hacked in Less than 72 Hours by Researchers Equipped with Claude, the AI of Anthropic
- [El Economista](https://eleconomista.com.mx/tecnologia/grupo-investigadores-logra-hackear-openai-utilizando-claude-20260918-834155.html) — A Group of Researchers Manages to Hack OpenAI Using Claude
- [El Mundo](https://elmundo.es/economia/empresas/2026/09/18/6aace6c0fdddff4d318b45a3.html) — Hackers Used Anthropic Claude to Enter OpenAI
- [Mediafax](https://mediafax.ro/tehnologie/cercetatori-in-securitate-au-demonstrat-folosind-claude-cum-pot-ajunge-la-codul-intern-al-openai-23809543) — Security Researchers Have Demonstrated, Using Claude, How They Can Reach OpenAI's Internal Code
- [dev.by](https://devby.io/news/issledovateli-vzlomali-openai-s-pomoschu-claude-i-poluchili-dostup-k-vnutrennemu-kodu) — Researchers Hacked OpenAI Using Claude and Accessed the Internal Code.
- [Business Insider](https://businessinsider.com/hacktron-ai-cybersecurity-startup-hack-openai-using-claude-2026-9) — This tiny cybersecurity startup managed to hack OpenAI using Claude
- [Startup Fortune](https://startupfortune.com/researchers-used-anthropics-claude-to-hack-into-openais-own-systems) — Researchers Used Anthropic's Claude to Hack Into OpenAI's Own Systems
- [Hola News](https://holanews.com/un-equipo-de-investigadores-jaqueo-a-openai-con-claude-la-ia-de-anthropic-segun-un-medio) — A Team of Researchers Checked OpenAI with Claude, the AI of Anthropic, According to a Media
- [OfficeChai](https://officechai.com/ai/hackers-were-able-to-use-claude-to-hack-openai-employee-accounts-and-create-a-pr-in-its-repo) — Hackers Were Able To Use Claude To Hack OpenAI Employee Accounts And Create A PR In Its Repo
- [computerworld.dk](https://computerworld.dk/art/296871/it-sikkerhedsfolk-broed-ind-i-openais-systemer-ved-hjaelp-af-anthropics-claude-her-er-fremgangsmaaden) — Cybersecurity Experts Broke Into OpenAI's Systems Using Anthropic's Claude: Here's How
- [Unknown](https://neakriti.gr/kosmos/2200550_espasan-tin-asfaleia-tis-openai-me-ergaleio-tis-anthropic-pos-apektisan-prosbasi-se) — They "Broke" OpenAI's Security with an Anthropic Tool - How They Gained Access to an Employee's Account
- [cybernoz.com](https://cybernoz.com/researchers-use-claude-opus-5-to-hack-openai-forum-and-reach-internal-repositories) — Researchers Use Claude Opus 5 to Hack OpenAI Forum and Reach Internal Repositories
- [National Cyber Security](https://nationalcybersecurity.com/openai-hacked-by-small-team-of-white-hat-security-researchers-using-anthropics-claude-opus-5-hacking-cybersecurity-infosec-comptia-pentest-hacker) — OpenAI hacked by small team of white hat security researchers using Anthropic's Claude Opus 5 | #hacking | #cybersecurity | #infosec | #comptia | #pentest | #hacker - National Cyber Security Consulting
- [Forbes](https://forbes.com/sites/siladityaray/2026/09/18/security-researchers-hacked-into-openai-using-anthropics-claude) — Security Researchers Hacked Into OpenAI Using Anthropic’s Claude
- [IT Security News - cybersecurity, infosecurity news](https://itsecuritynews.info/researchers-use-claude-opus-5-to-hack-openai-forum-and-reach-internal-repositories) — Researchers Use Claude Opus 5 to Hack OpenAI Forum and Reach Internal Repositories
- [BITRSS](https://bitrss.com/%E5%8D%8E%E5%B0%94%E8%A1%97%E6%97%A5%E6%8A%A5-%E9%BB%91%E5%AE%A2%E5%88%A9%E7%94%A8-anthropic-claude-%E5%85%A5%E4%BE%B5-openai-%E5%86%85%E9%83%A8%E7%B3%BB%E7%BB%9F-252359) — 华尔街日报：黑客利用 Anthropic Claude 入侵 OpenAI 内部系统
- [De Tijd](https://tijd.be/ondernemen/technologie/onderzoekers-kraken-chatgpt-account-met-anthropic-model/10686502.html) — Researchers Crack ChatGPT Account with Anthropic Model
- [Estado Red](https://estadored.mx/2026/09/claude-ayuda-a-hackear-openai) — Claude Helps Hack OpenAI
- [Dagens industri](https://di.se/live/utredare-anvande-anthropic-for-att-hacka-open-ai) — Investigators Used Anthropic to Hack Open AI
- [Cyber Security News](https://cybersecuritynews.com/opus-5-to-help-exploit-openai-flaws) — Researchers Use Claude Opus 5 to Hack OpenAI Forum and Reach Internal Repositories
- [vz.ru](https://vz.ru/news/2026/9/18/1452579.html) — The Hackers Hacked OpenAI Systems Using the Claude I.E. Model.
- [VentureBeat](https://venturebeat.com/security/openai-hacked-by-small-team-of-white-hat-security-researchers-using-anthropics-claude-opus-5) — OpenAI hacked by small team of white hat security researchers using Anthropic's Claude Opus 5
- [BizToc](https://biztoc.com/x/ffd998758909a248) — Hackers Used Anthropic’s Claude to Break Into OpenAI
- [Cryptocurrency News | Cryptocurrency Prices | Market Cap](https://cryptocurrency.com.tr/security-breach-exposes-openai-vulnerabilities-via-anthropics-claude-ai-wsj-381809) — Security breach exposes OpenAI vulnerabilities via Anthropic’s Claude AI: WSJ

## Questions

**How did researchers gain access to OpenAI's systems?**

Researchers from Hacktron AI exploited a vulnerability in OpenAI's community forum, specifically related to image processing. This allowed them to access an OpenAI employee's ChatGPT account, which was linked to the company's code repository, thereby gaining access to internal systems.

**Which AI model was used in the hack?**

The researchers used Anthropic's Claude AI model, specifically Claude Opus 5, to carry out the breach. Initial attempts with an earlier version, Claude Opus 4.8, were unsuccessful.

**What was the outcome for the researchers?**

OpenAI acknowledged the security flaws discovered by the Hacktron AI team and rewarded them with $6,500 as part of their bug bounty program. The company stated it has since patched the vulnerabilities.

---

MediaBias News — https://mediabias.news. Reproduced from https://mediabias.news/business/researchers-hack-openai-using-anthropic-ai-model. Please cite as: MediaBias News, "Researchers Hack OpenAI Using Anthropic AI Model", https://mediabias.news/business/researchers-hack-openai-using-anthropic-ai-model