Anthropic says its Claude models 'gained unauthorized access' to other organizations' systems

Skip NavigationMarketsPre-MarketsU.S. MarketsEurope MarketsChina MarketsAsia MarketsWorld MarketsCurrenciesPrediction MarketsCryptocurrencyFutures & CommoditiesBondsFunds & ETFsBusinessEconomyFinanceHealth & ScienceMediaReal EstateEnergyClimateTransportationInvestigationsIndustrialsRetailWealthSportsLifeSmall BusinessInvestingPersonal FinanceFintechFinancial AdvisorsOptions ActionETF StreetBuffett ArchiveEarningsTrader TalkTechCybersecurityAIEnterpriseInternetMediaMobileSocial MediaCNBC Disruptor 50Tech GuidePolitics & PolicyWhite HousePolicyDefenseCongressExpanding OpportunityEurope PoliticsChina PoliticsAsia PoliticsWorld PoliticsVideoLatest VideoFull EpisodesLivestreamTop VideoLive AudioEurope TVAsia TVCNBC PodcastsCEO InterviewsDigital OriginalsWatchlistInvesting ClubTrust PortfolioAnalysisTrade AlertsMeeting VideosHomestretchJim's ColumnsEducationSubscribePROPro NewsJosh BrownMike SantoliCalls of the DayMy PortfolioLivestreamFull EpisodesStock ScreenerMarket ForecastOptions InvestingChart InvestingSubscribeLivestreamMenuMake ItselectUSAINTLLivestreamSearch quotes, news & videosLivestreamWatchlistSIGN INCreate free accountMarketsBusinessInvestingTechPolitics & PolicyVideoWatchlistInvesting ClubPROLivestreamMenu (function() { const userMembershipOnLastVisit = localStorage.getItem("cnbcMeta__cachedUserMembership"); if (!userMembershipOnLastVisit) { // If no user membership is stored, user is likely not logged in. // These users will typically see the banner. return; } const { isPro, isInvestingClubUser } = JSON.parse(userMembershipOnLastVisit); if (!(isPro || isInvestingClubUser)) { // We’re rendering the top banner for users who are not pro or investing club members. return; } // Hide the banner for pro and investing club users. const style = document.createElement("style"); style.id = "top-banner-placeholder-style"; style.innerHTML = "[data-top-banner-placeholder] { display: none; }"; document.head.appendChild(style); })() TechAnthropic says its Claude models 'gained unauthorized access' to other organizations' systemsPublished Thu, Jul 30 20267:27 PM EDTUpdated 36 Min AgoAshley Capoot@/in/ashley-capoot/WATCH LIVEKey PointsAnthropic said it discovered three instances where its Claude AI models accessed the internet during an evaluation and "gained unauthorized access to the real systems of three different organizations." The company said it found these incidents after carrying out a "a large-scale retrospective review" of its cybersecurity evaluations. Anthropic said the review was prompted by a separate but similar security incident that OpenAI disclosed last week. In this articleANTHR.FGFollow your favorite stocksCREATE FREE ACCOUNTDario Amodei, co-founder and chief executive officer of Anthropic, at Bloomberg House during the World Economic Forum (WEF) in Davos, Switzerland, on Tuesday, Jan. 20, 2026. Chris Ratcliffe | Bloomberg | Getty ImagesAnthropic on Thursday said it discovered three instances where its Claude artificial intelligence models accessed the internet during an evaluation and "gained unauthorized access to the real systems of three different organizations." The company said it found these incidents after carrying out a "a large-scale retrospective review" of its cybersecurity evaluations. Anthropic said the review was prompted by a separate but similar security incident that OpenAI disclosed last week. OpenAI said a combination of its models escaped an isolated testing environment that had very limited internet access. The models chained together a series of vulnerabilities to reach the open web and eventually gain access to Hugging Face, which operates an open-source developer platform. In the three incidents that Anthropic detected, its models accessed the internet while i