<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>RuntimeRebel — #Data Leakage</title><description>Cybersecurity articles tagged #Data Leakage on RuntimeRebel.</description><link>https://runtimerebel.com</link><item><title>Rogue LLM Endpoints: Data Exposure &amp; RCE Risk for AI Agents</title><link>https://runtimerebel.com/blog/rogue-llm-endpoints-data-exposure-rce-risk-for-ai-agents</link><guid isPermaLink="true">https://runtimerebel.com/blog/rogue-llm-endpoints-data-exposure-rce-risk-for-ai-agents</guid><description>Unverified LLM endpoints pose significant risks, enabling data leakage and potential remote code execution via compromised AI agent sessions.</description><pubDate>Tue, 01 Sep 2026 02:48:52 GMT</pubDate><category>LLM Security</category><category>AI Agents</category><category>Honeypot</category><category>Data Leakage</category><category>Supply Chain Attack</category></item><item><title>LLM Persistent Memory and Contextual Integrity Risks</title><link>https://runtimerebel.com/blog/llm-persistent-memory-and-contextual-integrity-risks</link><guid isPermaLink="true">https://runtimerebel.com/blog/llm-persistent-memory-and-contextual-integrity-risks</guid><description>Analysis of new research on LLM contextual integrity, persistent memory risks, and how frontier models leak sensitive user data over time.</description><pubDate>Tue, 18 Aug 2026 16:24:01 GMT</pubDate><category>Artificial Intelligence</category><category>Data Leakage</category><category>Privacy</category><category>Zero-Day</category></item><item><title>LLM API Flaw Exposes Secrets in OpenAI, Anthropic, Google Traces</title><link>https://runtimerebel.com/blog/llm-api-flaw-exposes-secrets-in-openai-anthropic-google-traces</link><guid isPermaLink="true">https://runtimerebel.com/blog/llm-api-flaw-exposes-secrets-in-openai-anthropic-google-traces</guid><description>A flaw in OpenAI, Anthropic, and Google AI APIs allowed researchers to recover hidden reasoning, API keys, and passwords from exposed session logs.</description><pubDate>Thu, 13 Aug 2026 16:45:14 GMT</pubDate><category>OpenAI</category><category>Anthropic</category><category>Google</category><category>Data Leakage</category><category>AI Security</category></item><item><title>Financial Institutions Expose Customer Data via Cookie Trackers</title><link>https://runtimerebel.com/blog/financial-institutions-expose-customer-data-via-cookie-trackers</link><guid isPermaLink="true">https://runtimerebel.com/blog/financial-institutions-expose-customer-data-via-cookie-trackers</guid><description>European and US financial institutions inadvertently expose customer data to ad platforms through website cookie trackers, raising significant privacy and compliance…</description><pubDate>Wed, 22 Jul 2026 17:23:56 GMT</pubDate><category>Financial Institutions</category><category>Cookie Trackers</category><category>Data Leakage</category><category>Privacy</category><category>GDPR</category><category>CCPA</category><category>Ad Platforms</category></item><item><title>SASE AI Blind Spot: Why Packet Inspection Fails Modern Workflows</title><link>https://runtimerebel.com/blog/sase-ai-blind-spot-why-packet-inspection-fails-modern-workflows</link><guid isPermaLink="true">https://runtimerebel.com/blog/sase-ai-blind-spot-why-packet-inspection-fails-modern-workflows</guid><description>Enterprise data leakage via generative AI tools and browser extensions exposes critical flaws in traditional SASE models that rely solely on packet inspection.</description><pubDate>Wed, 15 Jul 2026 13:46:55 GMT</pubDate><category>SASE</category><category>GenAI</category><category>Data Leakage</category><category>Browser Security</category><category>DLP</category></item><item><title>Squidbleed: Heartbleed-Style Data Exposure in Squid Proxy</title><link>https://runtimerebel.com/blog/squidbleed-heartbleed-style-data-exposure-in-squid-proxy</link><guid isPermaLink="true">https://runtimerebel.com/blog/squidbleed-heartbleed-style-data-exposure-in-squid-proxy</guid><description>A critical flaw dubbed Squidbleed in Squid Proxy, affecting versions 3.5-6.x, enables Heartbleed-style memory leakage exposing user credentials and session data.</description><pubDate>Mon, 22 Jun 2026 17:38:08 GMT</pubDate><category>Squidbleed</category><category>Squid Proxy</category><category>Data Leakage</category><category>Proxy Vulnerability</category><category>Heartbleed Style</category><category>SQUID 2023 5</category><category>SQUID 2023 6</category><category>Memory Corruption</category></item><item><title>OpenClaw AI Agent Vulnerabilities: Code Execution &amp; Data Leakage</title><link>https://runtimerebel.com/blog/openclaw-ai-agent-vulnerabilities-code-execution-data-leakage</link><guid isPermaLink="true">https://runtimerebel.com/blog/openclaw-ai-agent-vulnerabilities-code-execution-data-leakage</guid><description>New research reveals OpenClaw AI agents can be tricked into executing malicious code or exfiltrating sensitive data via seemingly benign inputs like vCards and location…</description><pubDate>Thu, 11 Jun 2026 20:56:08 GMT</pubDate><category>OpenClaw</category><category>AI Agent</category><category>Code Execution</category><category>Data Leakage</category><category>Imperva</category><category>Varonis</category><category>Input Validation</category></item><item><title>Insecure Self-Hosted AI: 1 Million Exposed Services Risks Analyzed</title><link>https://runtimerebel.com/blog/insecure-self-hosted-ai-1-million-exposed-services-risks-analyzed</link><guid isPermaLink="true">https://runtimerebel.com/blog/insecure-self-hosted-ai-1-million-exposed-services-risks-analyzed</guid><description>A security scan of 1 million exposed AI services reveals critical vulnerabilities in self-hosted LLM infrastructure and misconfigured model deployments.</description><pubDate>Tue, 05 May 2026 12:35:57 GMT</pubDate><category>AI Security</category><category>LLM Exposure</category><category>Cloud Misconfiguration</category><category>Data Leakage</category></item><item><title>OpenSSL: Data Leakage &amp; DoS Vulnerabilities Patched</title><link>https://runtimerebel.com/blog/openssl-data-leakage-dos-vulnerabilities-patched</link><guid isPermaLink="true">https://runtimerebel.com/blog/openssl-data-leakage-dos-vulnerabilities-patched</guid><description>OpenSSL patches seven vulnerabilities, including a data leakage flaw and multiple denial-of-service risks. Update immediately to secure cryptographic communications.</description><pubDate>Thu, 09 Apr 2026 00:35:05 GMT</pubDate><category>OpenSSL</category><category>Data Leakage</category><category>DoS</category><category>Cryptography</category><category>TLS</category><category>SSL</category></item><item><title>Securing Shadow AI: How to Discover and Govern Unauthorized SaaS Tools</title><link>https://runtimerebel.com/blog/securing-shadow-ai-how-to-discover-and-govern-unauthorized-saas-tools</link><guid isPermaLink="true">https://runtimerebel.com/blog/securing-shadow-ai-how-to-discover-and-govern-unauthorized-saas-tools</guid><description>Learn how security teams can identify shadow AI usage, monitor OAuth grants, and implement governance to prevent corporate data leakage in SaaS environments.</description><pubDate>Mon, 16 Mar 2026 16:28:57 GMT</pubDate><category>Shadow AI</category><category>SaaS Security</category><category>Data Leakage</category><category>AI Governance</category><category>Oauth Security</category></item></channel></rss>