30
42
Amazon Kiro Prompt Injection Can Exfiltrate Sensitive Data Through Kiro Powers
Mindgard found a prompt injection flaw in Amazon Kiro IDE letting attacker-controlled workspace files exfiltrate sensitive local data; fixed in version 0.8.140.
Mindgard disclosed a prompt injection flaw in Amazon Kiro, an agentic AI IDE, that lets attacker-controlled repository content steer the agent into exfiltrating sensitive workspace data through Kiro Powers, which bundles MCP server configurations, POWER.md steering files, hooks, and contextual knowledge. Exploitation requires the user to open a malicious project via a workspace file and send any message to the agent; difficulty is rated low and it works in both trusted and untrusted workspaces. Amazon fixed the issue in Kiro IDE 0.8.140; the flaw has no CVE identifier and follows earlier Kiro bugs including CVE-2026-10591, plus related prompt-injection and code-execution issues in Codex CLI, Cursor, Gemini CLI, Copilot CLI, and Claude Code.
55
Manic Android Malware Exfiltrates Data From Offline Phones via Nearby Infected Devices
ThreatFabric identified Manic, an Android banking malware and spyware targeting Ukrainian and European financial apps with novel offline Wi-Fi mesh data exfiltration via nearby infected devices.
ThreatFabric reported a new Android malware family called Manic combining banking fraud and surveillance capabilities, targeting 169 package IDs across Ukrainian banks, government and identity services, messaging apps, and Russian and European financial institutions. The malware uses phishing sites and dropper apps impersonating utilities for distribution and relies on accessibility services and notification permissions for keylogging, overlays, and remote control. It introduces a store-and-forward relay mechanism that stages encrypted data locally and relays it through nearby infected devices via Wi-Fi Direct, Bluetooth RFCOMM, or BLE GATT, supporting up to four hops when the primary device lacks internet access. Activity dates back to February 2026, with active development through late July.
72
AI Flaws in Amazon Bedrock, LangSmith, and SGLang Enable Data Exfiltration and RCE
60
50
45
30
55
57
30
30
30
60
45
Mapping out your unknown: A threat hunter’s guide to GitHub
Datadog Security Labs publishes a threat-hunting guide with audit-log queries to detect GitHub token theft, device code phishing, and source code exfiltration.
Datadog's threat-hunting guide covers GitHub audit log queries for detecting compromised accounts, stolen personal access tokens, and malicious OAuth app authorizations. Attackers typically obtain credentials through phishing, credential stuffing, leaked secrets, or device code phishing, then map private repositories, exfiltrate source code, and pivot into connected cloud and CI/CD environments. The guide maps detections to MITRE techniques like T1078 and T1528 and documents GitHub logging quirks affecting attribution, token metadata, and visibility fields.
45
New AI Workflow Identity Hijacking Attack Lets Hackers Exfiltrate Sensitive Data
Noma Labs disclosed Workflow Identity Hijacking, an AI automation flaw letting anonymous users trigger privileged data exfiltration without prompt injection or stolen credentials.
Noma Labs researcher Sasi Levi described Workflow Identity Hijacking, where AI workflows process untrusted input from low-privileged or anonymous users but execute downstream actions with the workflow creator's elevated permissions, turning the pipeline into an unauthenticated proxy. Unlike prompt injection, the model is not tricked; the flaw is a missing authorization check between the requester and the privileged actions. Noma Labs also disclosed and helped fix a similar issue in Google Workflows, and linked the problem to the earlier GitLost research on GitHub Agentic Workflows. Recommended mitigations include per-user identity propagation, least-privilege service accounts and authorization checks before every downstream action.
55
CVE-2026-57866: Apache Impala: Secrets Exfiltration via SSRF
Apache Impala CVE-2026-57866 lets authenticated users abuse ai_generate_text() to exfiltrate secrets from configured Hadoop credential providers via SSRF.
A server-side request forgery affects Apache Impala versions 4.4.0 through 4.5.1. Authenticated users with permission to execute the ai_generate_text() function can exfiltrate secrets provided by credential providers configured via hadoop.security.credential.provider.path in core-site.xml. The attacker must know the secret's key name, and Apache rates the issue 'important'.
33
Threat Brief: Maze Ransomware
Unit 42 issued a threat brief on Maze ransomware, noting an uptick across industries and exploitation of Pulse VPN and IE flaws for initial access.
Palo Alto Networks detected an uptick in Maze ransomware samples across finance, healthcare, government, and other sectors. Maze, a variant of ChaCha ransomware first seen in May 2019, is distributed via weaponized Word/Excel attachments and the Spelevo exploit kit leveraging CVE-2018-15982 and CVE-2018-4878, and has used CVE-2019-11510 (Pulse VPN) and CVE-2018-8174 (Internet Explorer). Operators establish a foothold, escalate privileges, move laterally, and exfiltrate files before encryption for extortion leverage.
55
Malicious MCP Servers Can Split Instructions to Make AI Coding Agents Exfiltrate Secrets
ASSET Research Group's GhostSplice technique splits malicious instructions across MCP channels, tricking AI coding agents into exfiltrating SSH keys, source code, and secrets.
ASSET Research Group disclosed GhostSplice, a prompt-injection technique in which a malicious Model Context Protocol (MCP) server splits an exfiltration instruction across a tool description and a tool result so no single fragment appears harmful. In the reference implementation, a benign-looking integrity_checker tool with fields alpha through delta is later paired with a project-scan result mapping those fields to .ssh/id_rsa, proprietary source, customers.csv, and .env. Tests across eleven API-tested models showed average compliance rising from 42% to 82% when instructions were split in two, with GPT-4o, Gemini 2.0 Flash, and Llama 3.3 70B going from 0% to 100%. The findings come from controlled lab tests, not a reported real-world intrusion, and no CVE identifiers had been assigned as of August 10, 2026.
63
42
60
45
30
55
47
60
60
30
55
60
42
60
57
57
42
55
30
30
45