Changes made to a running process, such as writing data into memory, modifying execution behavior, or injecting code into an existing process. Adversaries frequently modify processes to execute malicious payloads, evade detection, or gain escalated privileges.
Developer or CI invokes package managers/compilers (apt/yum + build-essential, npm/yarn/pnpm, pip/pip3, gem, cargo, go, maven/gradle). These write executable or script files into PATH or project dirs and immediately execute embedded lifecycle hooks (preinstall/postinstall, setup.py, npm scripts) that spawn shells or curl/wget, followed by egress to unfamiliar registries or domains.
Correlates ELF file execution with high-entropy writable memory segments and self-modifying code patterns.
Detection of packed Mach-O binaries unpacking into memory and transferring control to dynamically modified code segments.
Disabling or modifying the Linux Audit system through process termination (auditd killed), service management (systemctl stop auditd), or tampering with rule/configuration files (/etc/audit/audit.rules, audit.conf). Defender view: suspicious execution of auditctl/systemctl commands, file modifications to audit rules, or sudden absence of audit logs correlated with privileged execution.
Detection of encoded payloads being decoded and executed in-memory using scripting tools or third-party decoders.
Detects malicious injection behavior involving memory allocation, remote thread queuing via APC (e.g., QueueUserAPC), and altered thread context within another live process to execute unauthorized code under legitimate context.
Detects PE injection through a behavioral sequence where one process opens (OpenProcess) a handle to another, allocates remote memory (VirtualAllocEx), writes a PE header (MZ) or shellcode (WriteProcessMemory), then initiates a new thread (CreateRemoteThread or NtCreateThreadEx) in that process—executing injected code in memory without touching disk. Optional: injects a trampoline or shellcode that unpacks/reflectively maps the payload.
Detects credential harvesting via userland API hooking (e.g., SetWindowsHookEx, IAT, or inline patching) by correlating memory modifications with hook installation functions and suspicious module loads in credential-sensitive processes like lsass.exe, explorer.exe, or winlogon.exe.
Detects DYLD_INSERT_LIBRARIES abuse to hook credential-sensitive applications by correlating process spawns with unauthorized library injection and monitoring changes to the __TEXT segment (code) of credential handling binaries.
Detects adversary behavior where the command-line arguments of a running process are overwritten in memory to spoof the process name, typically replacing it with a benign or misleading string. The detection correlates unexpected null byte sequences, discrepancies between `/proc/<pid>/cmdline` and process ancestry, and suspicious memory writes shortly after process start.
Correlated evidence of anomalous browser/network behavior (suspicious external resource fetches and script injection patterns) followed by atypical child processes, ephemeral execution contexts, memory modification or process injection, and unexpected file drops. Defender sees network requests to previously unseen/suspicious domains or resources + browser process spawning unusual children or loading unsigned modules + file writes or registry changes shortly after those requests.
Correlated evidence where Safari/Chrome/WebKit-based processes issue network requests for uncommon or obfuscated JS resources followed by spawning of script interpreters, launchd or ad-hoc binaries, unusual child processes, or dynamic library loads into browser processes. Defender sees: proxy/HTTP logs with suspicious resource content + unifiedlogs/ASL showing browser/plugin crashes or extension loads + process events indicating child process creation and file writes to /var/folders or /tmp shortly after the fetch.
Detection of anti-malware quarantining or flagging a tool, followed by a new binary written to disk with a similar function or name and a resumed process chain.
Detects hijacking of an existing thread (OpenThread) through a behavioral chain involving thread suspension (SuspendThread), memory modification (VirtualAllocEx + WriteProcessMemory), context manipulation (SetThreadContext), and thread resumption—all within another live process's address space (ResumeThread).
Detects ELF binaries written to disk that demonstrate anomalous file size or entropy, quickly followed by execution or memory region writes into remote processes (e.g., using ptrace).
Identifies Mach-O binaries dropped into temporary directories with abnormally high binary size or padding patterns, followed by privilege escalation, `exec`, or memory mapping of other processes.
Tracks modification of executables or interpreter payloads (e.g., Mach-O, dylib) that mutate across runs—using scripting engines, JIT compilers, or side-loaded plugins.
Detects the use of message-based injection by monitoring for sequences involving FindWindow (EnumWindows or EnumChildWindows), VirtualAllocEx or related API calls, combined with suspicious PostMessage/SendMessage (e.g., LVM_SETITEMPOSITION) use to SysListView32 controls, followed by LVM_SORTITEMS invocation instead of WriteProcessMemory.
Detects renamed binaries or scripts placed into trusted paths like /usr/bin or /lib with mismatched metadata or unexpected creation/modification times.
Detects adversary use of suspended process creation, using the CREATE_SUSPENDED flag via CreateProcess, followed by unmapping the memory of the child process (NtUnmapViewOfSection) and replacing it with malicious code via VirtualAllocEx/WriteProcessMemory, then SetThreadContext and ResumeThread to begin execution within the hollowed process.
Detects custom archive routines by correlating script execution (Python, Perl, Bash) with creation of high-entropy files in temporary or user directories. Flags processes performing unusual bitwise operations or writing files without standard compression headers.
Detects custom archiving by monitoring execution of Swift/Objective-C apps or scripts producing high-entropy files with non-standard headers. Correlates unified logs of abnormal NSFileHandle/NSData operations, memory use of XOR/bitwise operations, and file creation events.
Detects the redirection of syscall execution flow via modification of VDSO code stubs or GOT entries to load and execute a malicious shared object through mmap and ptrace.
Detects thread local storage (TLS) callback injection by monitoring memory modifications to PE headers and TLS directory structures during or after process hollowing events, followed by anomalous thread behavior prior to main entry point execution.
Adversary gains high integrity or special privileges (e.g., SeDebugPrivilege), locates a running browser process, opens it with write/inject rights, and modifies it (e.g., CreateRemoteThread / DLL load) to inherit cookies/tokens or establish a browser pivot. Optional step: create a new logon session or use explicit credentials, then drive the victim browser to intranet resources.
Detects ptrace- or memfd-based process injection through audit logs capturing system calls (e.g., ptrace, mmap) targeting running processes along with suspicious file descriptors or memory writes.
1) Package manager or curl/wget installs/upgrades from non-approved repos or unsigned packages; 2) new ELF written into PATH directories or replacement of existing binaries/libraries; 3) first run leads to unexpected child processes or outbound connections.
MSBuild.exe is invoked outside expected developer/build contexts or with anomalous arguments (e.g., non-canonical paths, remote shares, Base64/obfuscated property values). Within a short window, it (a) spawns high-risk LOLBins/script interpreters, (b) writes new PE/DLL/script artifacts into user-writable paths and executes them, (c) loads unsigned/user-writable modules, (d) performs memory injection/thread creation into other processes, and/or (e) initiates outbound network connections.
Windows environmental validation behavioral chain: (1) Rapid system discovery reconnaissance through WMI queries, registry enumeration, and network share discovery, (2) Environment-specific artifact collection (hostname, domain, IP addresses, installed software, hardware identifiers), (3) Cryptographic operations or conditional logic based on collected environmental values, (4) Selective payload execution contingent on environmental validation results, (5) Temporal correlation between discovery activities and subsequent execution or network communication