[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$frmth6Xj96MlaNbKURa-BQragSOd9aNxivMVXTiFiDDI":3},{"article":4,"iocs":37,"watch_terms":56,"solutions":57,"lesson":10},{"id":5,"title":6,"slug":7,"summary":8,"ai_summary":9,"brief":10,"url":11,"image_url":12,"published_at":13,"ingested_at":14,"relevance_score":15,"entities":16,"category_id":17,"category":18,"article_tags":21},"79a2f2f1-034e-4274-b16f-0b05f1805f97","AI Flaws in Amazon Bedrock, LangSmith, and SGLang Enable Data Exfiltration and RCE","ai-flaws-in-amazon-bedrock-langsmith-and-sglang-enable-data-exfiltration-and-rce","Cybersecurity researchers have disclosed details of a new method for exfiltrating sensitive data from artificial intelligence (AI) code execution environments using domain name system (DNS) queries. In a report published Monday, BeyondTrust revealed that Amazon Bedrock AgentCore Code Interpreter's sandbox mode permits outbound DNS queries that an attacker can exploit to enable interactive shells","Security researchers disclosed critical flaws in Amazon Bedrock, LangSmith, and SGLang that enable data exfiltration and remote code execution through DNS tunneling, token theft, and unsafe deserialization. Amazon Bedrock's sandbox mode allows outbound DNS queries despite network isolation claims (CVSS 7.5), LangSmith suffers from URL parameter injection enabling account takeover (CVE-2026-25750, CVSS 8.5), and SGLang contains unauthenticated RCE vulnerabilities via unsafe pickle deserialization (CVE-2026-3059\u002F3060, CVSS 9.8). Organizations are urged to migrate to VPC mode, implement DNS firewalls, audit IAM permissions, and patch immediately.",null,"https:\u002F\u002Fthehackernews.com\u002F2026\u002F03\u002Fai-flaws-in-amazon-bedrock-langsmith.html","https:\u002F\u002Fblogger.googleusercontent.com\u002Fimg\u002Fb\u002FR29vZ2xl\u002FAVvXsEhlC6FiLJ9YVGC1W0eo-MmFsPTu2DNqMSdo-QKnv1gdH_HpaKV3zPaWZrQTGNdklpv62BXb3ECiBqlkR1BzLbfz0tFWfMKNM1vZq88yf90XpycB2OSDq3NScWav6ZO_4IVjCWaJRJLwvthFo-7VJ-Uc8qijyycpXfkQRHcPr9pf-QVgGzyOeBFnwfOGnmLN\u002Fs1600\u002Flang-ai.jpg","2026-03-17T16:39:00+00:00","2026-03-17T19:00:10.685629+00:00",9,[],"839da5c1-3c34-47e2-9499-f7201640e3ac",{"id":17,"icon":10,"name":19,"slug":20},"AI Security","ai-security",[22,27,32],{"category":23},{"id":24,"icon":10,"name":25,"slug":26},"574f766a-fb3f-487c-8d2c-0720ae75471b","Zero-day","zero-day",{"category":28},{"id":29,"icon":10,"name":30,"slug":31},"80544778-fabb-4dcd-aa35-17492e5dcf4f","Vulnerabilities","vulnerabilities",{"category":33},{"id":34,"icon":10,"name":35,"slug":36},"c70f3a41-2f0c-4608-870d-b8cbcd8be076","Cloud Security","cloud-security",[38,42,45,48,52],{"type":39,"value":40,"context":41},"cve","CVE-2026-25750","LangSmith URL parameter injection vulnerability enabling token theft and account takeover (CVSS 8.5)",{"type":39,"value":43,"context":44},"CVE-2026-3059","SGLang unauthenticated RCE via unsafe pickle deserialization in ZeroMQ broker (CVSS 9.8)",{"type":39,"value":46,"context":47},"CVE-2026-3060","SGLang unsafe pickle deserialization vulnerability (CVSS 9.8)",{"type":49,"value":50,"context":51},"malware","DNS tunneling C2 channel","Amazon Bedrock AgentCore Code Interpreter abuse for command-and-control via DNS queries",{"type":53,"value":54,"context":55},"domain","smith.langchain[.]com","LangSmith cloud deployment vulnerable to parameter injection attacks",[],{"controls":58,"vendors":141},[59,87,116],{"slug":60,"name":61,"problem":62,"products":63},"privileged-access-management","Privileged access management (PAM)","Admin accounts stop being one stolen password away from control of everything.",[64,72,79],{"vendor_slug":65,"vendor_name":66,"vendor_logo":67,"product_slug":68,"product_name":69,"url":70,"description":71},"wallix","WALLIX","\u002Fvendor-logos\u002Fwallix.png","wallix-pam","WALLIX PAM","https:\u002F\u002Fwww.wallix.com\u002Fproducts\u002Fprivileged-access-management-software\u002F","Privileged access management with session recording, for IT and OT environments.",{"vendor_slug":73,"vendor_name":74,"vendor_logo":75,"product_slug":60,"product_name":76,"url":77,"description":78},"threatlocker","ThreatLocker","\u002Fvendor-logos\u002Fthreatlocker.png","Privileged Access Management","https:\u002F\u002Fwww.threatlocker.com\u002Fcapabilities\u002Fprivileged-access-management","Gives elevated rights to specific applications only where needed, so users do not need local admin accounts.",{"vendor_slug":80,"vendor_name":81,"vendor_logo":82,"product_slug":83,"product_name":84,"url":85,"description":86},"delinea","Delinea","\u002Fvendor-logos\u002Fdelinea.png","secret-server","Delinea Secret Server","https:\u002F\u002Fdelinea.com\u002Fproducts\u002Fsecret-server","Vault for privileged credentials with discovery, rotation and session recording.",{"slug":88,"name":89,"problem":90,"products":91},"cloud-security-posture","Cloud security posture management (CSPM and CNAPP)","Misconfigured cloud storage and over-permissive roles get caught before they leak.",[92,100,108],{"vendor_slug":93,"vendor_name":94,"vendor_logo":95,"product_slug":96,"product_name":97,"url":98,"description":99},"crowdstrike","CrowdStrike","\u002Fvendor-logos\u002Fcrowdstrike.png","falcon-cloud-security","Falcon Cloud Security","https:\u002F\u002Fwww.crowdstrike.com\u002Fen-us\u002Fplatform\u002Fcloud-security\u002F","Cloud posture management and workload protection.",{"vendor_slug":101,"vendor_name":102,"vendor_logo":103,"product_slug":104,"product_name":105,"url":106,"description":107},"wiz","Wiz","\u002Fvendor-logos\u002Fwiz.png","wiz-cloud","Wiz Cloud","https:\u002F\u002Fwww.wiz.io\u002Fplatform\u002Fwiz-cloud","Agentless cloud security posture and attack path analysis across major cloud providers.",{"vendor_slug":109,"vendor_name":110,"vendor_logo":111,"product_slug":112,"product_name":113,"url":114,"description":115},"palo-alto-networks","Palo Alto Networks","\u002Fvendor-logos\u002Fpalo-alto-networks.png","cortex-cloud","Cortex Cloud","https:\u002F\u002Fwww.paloaltonetworks.com\u002Fcortex\u002Fcloud","Cloud security platform covering posture management and runtime protection.",{"slug":20,"name":117,"problem":118,"products":119},"AI application security","Prompt injection, data leaks through AI tools and unapproved AI use get caught.",[120,125,133],{"vendor_slug":109,"vendor_name":110,"vendor_logo":111,"product_slug":121,"product_name":122,"url":123,"description":124},"prisma-airs","Prisma AIRS","https:\u002F\u002Fwww.paloaltonetworks.com\u002Fai-security\u002Fprisma-airs","Security for AI applications, models and agents, including runtime protection against prompt injection.",{"vendor_slug":126,"vendor_name":127,"vendor_logo":128,"product_slug":129,"product_name":130,"url":131,"description":132},"boomi","Boomi","\u002Fvendor-logos\u002Fboomi.png","boomi-agentstudio","Boomi Agentstudio","https:\u002F\u002Fboomi.com\u002Fplatform\u002Fagentstudio\u002F","Agent builder whose Agent Control Tower registers and monitors Boomi agents and agents on Amazon Bedrock, Microsoft Copilot and other platforms, keeps audit logs and can switch agents off where the provider allows it.",{"vendor_slug":134,"vendor_name":135,"vendor_logo":136,"product_slug":137,"product_name":138,"url":139,"description":140},"hiddenlayer","HiddenLayer","\u002Fvendor-logos\u002Fhiddenlayer.png","hiddenlayer-ai-security","HiddenLayer AI Security Platform","https:\u002F\u002Fhiddenlayer.com\u002F","Model scanning, AI red teaming and runtime protection for AI applications.",[142],{"slug":143,"name":144,"logo_path":145,"roles":146},"beyondtrust","BeyondTrust","\u002Fvendor-logos\u002Fbeyondtrust.png",[147],"reported_by"]