A quiet new class of malware is muscling into the spotlight and it is not smashing systems or tripping alarms. It slips in, blends with normal activity, and waits. In its latest Red Report, Picus Labs says attackers are pivoting from smash-and-grab ransomware to parasitic sleeperware that lives off compromised environments and monetizes access through stealthy data theft and extortion.
Ransomware Falls As Stealth Extortion Rises
Picus Labs analyzed more than one million malicious files and 15 million adversarial actions to rank real-world techniques using the MITRE ATT&CK framework. One signal stands out: the technique Data Encrypted for Impact (T1486) dropped from 21.00% of samples to 12.94%, a 38% relative decrease. Encryption-led shakedowns are giving way to long-term exploitation and data leverage.

Instead of detonating a payload that stops business cold, intruders increasingly choose to inhabit networks, accumulate credentials, quietly siphon valuable records, and later pressure victims with proof of access. Encryption still appears, but often as a secondary tactic rather than the opening move, echoing warnings from incident responders at NCC Group about a broader shift toward data-centric blackmail.
How Sleeperware Evades Detection Across Environments
Picus Labs again ranks Process Injection (T1055) as the most prevalent technique, underscoring the priority attackers place on dwell time. By injecting into legitimate processes, malware inherits trust, blends with routine memory activity, and avoids crude signatures. Command and Scripting Interpreter (T1059) and Credentials from Password Stores (T1555) follow, arming intruders with automation and authorized access.
The fastest climber is Virtualization/Sandbox Evasion (T1497), now sitting fourth. Modern samples detect analysis environments through timing checks, artifact scanning, and user interaction patterns, then refuse to execute under scrutiny. The result is a dangerous false negative: files that sail through automated gateways only to activate in production.
What makes this class “sleeperware” is patience paired with context awareness. Operators leverage stolen tokens, single sign-on trust, and cloud APIs. They harvest small data batches that look like routine traffic, often using living-off-the-land binaries and scripts to minimize noisy binaries on disk. The objective is survival, not spectacle.
Why Lurking Pays Better Than Locking For Attackers
Economics drive the change. Total network takeover carries risk: outages trigger rapid IR mobilization, media attention, and law-enforcement heat. By staying operationally invisible, adversaries can quietly exploit multiple revenue streams—exfiltrating regulated data, selling access to other crews, and staging repeated extortion rounds with escalating proof of compromise.
This approach also exploits security incentives. Many programs are tuned to catch break-ins and big anomalies, not to flag a known admin account copying “just enough” records after hours or a trusted process initiating memory injection. When detection hinges on loud signatures, patient intruders win by going quiet.

Tactics To Spot And Stop Sleeperware In Your Network
Picus Labs recommends tuning defenses to the dominant techniques rather than last year’s headlines. Start with memory-focused detection: monitor for Process Injection (T1055) indicators, such as suspicious thread creation across process boundaries and anomalous module loads, and ensure EDR policies are not neutered by whitelisting “too-trusted” binaries.
Harden identity and secrets. Lock down credential stores (T1555) with strict role-based access, password vault auditing, and hardware-backed key protection. Rotate access tokens aggressively, enforce step-up authentication for sensitive actions, and limit cloud API permissions to the minimum viable set to shrink the blast radius.
Counter sandbox-aware malware by diversifying analysis. Blend automated detonation with human-in-the-loop triage, randomized environment artifacts, and delayed execution windows to flush out time-bomb behaviors. Treat “no execute” outcomes as a signal, not a pass.
Shift telemetry to catch low-and-slow exfiltration. Baseline normal data egress by user and app, then alert on subtle deviations over longer windows. Plant canary records to generate high-fidelity signals if touched. Inspect command-and-scripting activity (T1059) for unusual parent-child process chains and off-hour spikes.
Resilience still matters. Even as encryption declines, immutable and isolated backups remain critical against destructive wipers and fallback ransomware. Test recovery paths regularly and segment backup networks to prevent token reuse from turning safety nets into stepping stones.
What Security Leaders Should Watch Next In Threat Trends
Expect sleeperware to deepen roots in cloud, identity, and supply chain pathways where trust is abundant and noise is cheap to mimic. Track the prevalence of T1497 evasion and T1055 injection in your own detections, not just in vendor reports, and recalibrate KPIs away from “ransomware blocked” toward “dwell time reduced” and “credential misuse detected.”
The bottom line from the Red Report is unambiguous: the most profitable attack is the one you do not notice. If security hinges on catching the crash, you will miss the parasite. Shift strategy to expose the quiet behaviors that sleeperware relies on, and you turn patience from the attacker’s advantage into your early warning.
