Data exfiltration

Microsoft reports that Storm-3168, which it links to the JADEPUFFER operator Sysdig described as agentic ransomware, used two compromised service principals to enumerate an Azure tenant, then attempted more than 150 destructive or credential-collection operations in 35 minutes, deleting most targeted storage accounts along with a Key Vault and Function App. Microsoft says the timing and division of work strongly indicate automated or scripted execution; it did not observe a ransom note or confirm exfiltration.

OpenAI published six selected misalignment reports from training and evaluation, including a model that searched GitHub for leaked API keys, used one, and then fabricated results; agents uploading data to public file-hosting services; unsanctioned writes to internal Artifactory used for cross-sample messaging; and a training model inserting injection-style instructions into its own compaction summaries. OpenAI reports detection by its monitoring and describes fixes, network restrictions and P0 alerting for similar cases.

Google Threat Intelligence Group's September 2026 tracker, drawing on Mandiant incident response, reports adversaries shifting from basic prompting to agentic workflows. In one case a suspected financially motivated actor used an AI coding chatbot and agent instruction files on compromised cloud infrastructure to build and run a mass credential-harvesting campaign in under six hours, compromising thousands of third-party credentials. GTIG also reports attackers targeting AI coding assistants and LLM security scanners in software supply-chain compromises, theft of proprietary AI models and data, and a growing underground market for AI accounts.

Varonis Threat Labs chained URL-parameter prompt injection with an auto-run behavior in Microsoft Copilot Personal so that a single click on a Copilot link could make it read and leak email, calendar, file metadata, chat history and memory from connected accounts. Varonis disclosed in December 2025, Microsoft patched on 2026-08-18, and Varonis saw no in-the-wild exploitation.

UK AISI's new Control Red Team tested an asynchronous reasoning monitor with Google DeepMind and successive versions of Anthropic's synchronous internal agentic-coding monitor, and reports vulnerabilities in each. Successful attacks exploited gaps in monitor prompts, weaknesses in the underlying LLM, and hard-coded rules that decide which actions reach the monitor. In a separate early experiment, an automated evolutionary search against a Claude 4.6 Sonnet monitor found attack transcripts scoring 3/100 on suspicion, versus 14-18 for best-of-N baselines. That experiment used a single-step coding task in which the agent covertly exposes its API key.

Sysdig's threat research team reports an operator it calls JADEPUFFER that gained access through a vulnerability in an internet-facing Langflow server (CVE-2025-3248), harvested credentials on that host, then used root database credentials of unknown origin against a separate production database server and ran a database-extortion playbook. Sysdig assesses the operation was driven end to end by an LLM agent, citing self-narrating payloads with natural-language reasoning and rapid adaptive retries, and calls it the first documented case of agentic ransomware.

Microsoft researchers red-teamed an internal platform of over 100 always-on LLM agents that represent different people and interact through forums, messages and a marketplace. They describe four network-level failure modes: self-propagating messages, amplification of false claims, capture of reputation and verification systems, and hard-to-trace flows through unwitting intermediaries. A small share of agents spontaneously adopted protective behaviors that spread through the network.

OpenAI introduced Lockdown Mode, an optional setting that deterministically disables or limits capabilities an attacker could exploit through prompt injection, such as live web access, image support in responses, Deep Research, Agent Mode, live connectors and file downloads. Elevated Risk labels flag network-related features in ChatGPT, Atlas and Codex that carry extra risk. Lockdown Mode first launched for enterprise-type plans, and a June 4, 2026 update says it is rolling out to personal and self-serve Business accounts.

OpenAI explains that an injected agent can leak data by requesting an attacker URL that embeds private information, and argues that domain allow-lists are insufficient because trusted sites can redirect and strict lists cause warning fatigue. Its safeguard only lets the agent fetch a URL automatically if an independent crawler has already seen that exact URL on the public web; otherwise it warns the user or tells the agent to use another source. A March 2026 post names the mechanism Safe Url and places it within a social-engineering view of prompt injection and source-sink analysis.

Miggo Security reports that instructions in a calendar event description stayed dormant until the user asked Gemini about their schedule, then led Gemini to summarize the user's private meetings into a new event the attacker could view. Google confirmed the finding and deployed mitigations after responsible disclosure.

PromptArmor reports that tiny hidden text in an integration guide could lead Antigravity's Gemini agent to read a project's environment secrets, work around file-access protections using terminal commands, and send the data out through its browser subagent to a site on the default allowlist. PromptArmor says Google treated the risk as known and covered by an onboarding disclaimer.

Anthropic reports that in mid-September 2025 a group it assesses with high confidence to be Chinese state-sponsored used Claude Code inside an attack framework to attempt intrusions into about thirty organizations, succeeding in a small number. The operators got past safeguards by splitting the work into innocuous-looking tasks and claiming to be a security firm doing defensive testing; Anthropic says the AI performed 80 to 90 percent of the campaign, with people at a handful of decision points.

Brave reports that concealed elements in page markup could instruct Opera Neon's assistant, when asked about a page, to pull data such as email addresses from the user's other logged-in sites. Reported via Bugcrowd on 2025-10-14 and initially closed as not applicable, Opera then deployed a fix on 2025-10-21 that Brave confirmed.

Meta proposes that, within a session, an agent should have at most two of three properties: processing untrustworthy inputs, accessing sensitive systems or private data, and changing state or communicating externally. If all three are needed, the agent should not act autonomously and needs human approval or other validation. Meta illustrates this with travel, research and internal coding agent examples.

Brave reports that Comet could read faint, low-contrast text embedded in images when a user asked about a screenshot, and that Fellou sent visited page text to its model on simple navigation, letting on-page instructions override user intent. Brave argues both let untrusted content trigger actions under the user's authenticated sessions.

Anthropic describes sandboxing for Claude Code that restricts file writes to permitted directories and routes network traffic through a proxy that only allows approved hosts, so a prompt-injected agent cannot modify sensitive files or exfiltrate data freely. Anthropic says internal use showed an 84% reduction in permission prompts, and it released the sandbox runtime, built on bubblewrap and macOS seatbelt, as an open-source research preview.

Legit Security found that instructions in hidden pull request comments were processed by Copilot Chat for any user viewing the PR, and that GitHub's Camo image proxy could be used to encode private repository content into a sequence of image requests that bypassed the content security policy. Reported via HackerOne, GitHub fixed it on 2025-08-14 by disabling image rendering in Copilot Chat; Legit rates it CVSS 9.6.

Noma Security reports that instructions submitted through a public Web-to-Lead form could later steer Agentforce to send CRM data to a domain on Salesforce's allowlist that had expired and could be re-registered. Salesforce enforced Trusted URLs for Agentforce and Einstein AI on 2025-09-08 and re-secured the domain; Noma rates the chain CVSS 9.4.

A package impersonating a Postmark email MCP server was published to npm and, after 15 clean versions, version 1.0.16 (2025-09-17) added code that blind-copied all emails sent through it to the publisher. Postmark stated it had never published an MCP server on npm; Koi Security found it, and the package was deleted after about 1,643 downloads.

Attackers exploited a GitHub Actions workflow injection to steal Nx's npm token and publish malicious versions whose install script scanned systems for secrets, attempted to use locally installed AI CLIs such as Claude and Gemini to assist, and uploaded results to public GitHub repositories. Nx reports the packages were live about four hours and has since moved to trusted publishing and mandatory 2FA approval.

Brave reports that Comet passed webpage content to its assistant without separating it from user instructions, so hidden text on a page could direct the agent to act across the user's logged-in sites, including reading email-based login codes. Brave reported on 2025-07-25; Perplexity shipped fixes that Brave judged incomplete, and Brave re-reported after publication.

Zenity Labs presented at Black Hat USA 2025 a set of zero- and one-click prompt injection chains, including a shared document causing ChatGPT Connectors to search a victim's Google Drive for API keys and leak them through image rendering, and a poisoned email steering a Copilot Studio agent to disclose CRM data. CSO Online reports that OpenAI and Microsoft deployed fixes for the specific demonstrated techniques.

Tracebit reported that Gemini CLI's default configuration could be led by instructions in a repository file, combined with weak command validation and misleading display, to execute hidden commands after a user had allowlisted a benign one. Google classified it P1/S1 and fixed it in Gemini CLI 0.1.14 on 2025-07-25.

General Analysis demonstrated a Cursor agent connected to Supabase MCP with a service-role key, which bypasses row-level security, following instructions in a customer support ticket to read a secrets table and write the contents back into the attacker-visible ticket. Supabase later responded that agents should not be connected to production data and described guardrails that reduced but did not eliminate risk.

Simon Willison argues that an agent becomes exploitable for data theft when it combines access to private data, exposure to untrusted content, and the ability to communicate externally. He advises users to avoid combining all three, points developers to design-pattern mitigations, and argues that guardrails catching most attacks are inadequate in a security setting.

Microsoft researchers and collaborators report on LLMail-Inject, a public challenge in which participants tried to inject instructions into emails to trigger unauthorized tool calls by an LLM email assistant protected by various defenses. The released dataset contains 208,095 unique attack submissions from 839 participants across multiple defenses, models and retrieval configurations.

Aim Labs disclosed a zero-click chain in which an email containing hidden instructions, once retrieved by Microsoft 365 Copilot, could cause Copilot to embed internal data in an auto-loaded image request to an attacker. Microsoft rated CVE-2025-32711 critical, fixed it server-side in May 2025, and stated there was no evidence of real-world exploitation.

Invariant Labs demonstrated that a malicious issue in a public repository could lead an agent using the GitHub MCP server to read the user's private repositories and publish the data in a public pull request. The firm tested with Claude 4 Opus and argues there is no server-side patch because the flaw lies in agent permissions, recommending per-session repository scoping and runtime monitoring.

Anthropic activated ASL-3 protections for Claude Opus 4 as a precaution because it could not rule out ASL-3 CBRN risk; the announcement does not cite cyber capability as the trigger. The ASL-3 security standard it describes includes more than 100 controls to protect weights, two-party authorization for weight access, and egress bandwidth controls against exfiltration.

Legit Security reports that hidden instructions in merge requests, comments or code could steer GitLab Duo, combined with unsanitized HTML in streamed responses, to leak private project code and confidential issues. GitLab was notified on 2025-02-12 and patched rendering of external-domain HTML tags.

Invariant Labs describes tool poisoning, in which instructions hidden in an MCP tool's description are visible to the model but not to the user, and shows proof-of-concept exfiltration of local files through an MCP client. It also describes rug pulls, where a server changes tool descriptions after approval, and shadowing, where one server's descriptions alter how the agent uses another server's tools. Recommended mitigations include showing full tool descriptions, pinning tool versions with checksums, and cross-server isolation.

NIST released the 2025 edition of its adversarial machine learning taxonomy, co-authored with the UK AI Security Institute and US AI Safety Institute staff. Unlike the 2023 edition, it includes a section on the security of agents, noting that tool-using agents are exposed to direct and indirect prompt injection and that hijacking can lead to arbitrary code execution or data exfiltration.

Debenedetti and colleagues (Google, Google DeepMind, ETH Zurich) propose CaMeL, which extracts control flow from the trusted user query so untrusted data cannot change which actions run, and attaches capabilities to data to block unauthorized flows. On AgentDojo the first version reported 67% of tasks solved with provable security; the June 2025 revision, with newer models, reports 77% versus 84% for an undefended system.