New Perseus Android banking malware monitors notes apps to steal sensitive data

March 20, 20268 min read6 sources
Share:
New Perseus Android banking malware monitors notes apps to steal sensitive data

Background and context

Researchers have disclosed a new Android banking malware family called Perseus that is being used in active campaigns to enable device takeover and financial fraud, according to reporting by The Hacker News. The malware is described as building on the foundations of the older Cerberus and Phoenix families, two names that have long been associated with Android banking trojans that abuse accessibility features, overlays, and remote control functions to steal credentials and defeat authentication flows (The Hacker News).

That lineage matters. Cerberus helped define a model for Android banking malware that mixed credential theft with on-device fraud support. Phoenix and other descendants expanded that playbook with improved loaders, modular components, and better evasion. Perseus appears to continue that trend: less a wholly new invention than a refinement of a proven criminal toolkit, adapted for current Android fraud operations.

The most notable detail in the latest reporting is Perseus’ ability to monitor notes applications for sensitive information. That is a meaningful shift in targeting. Banking malware has traditionally focused on banking apps, SMS messages, and notifications. By watching notes apps, operators can reach a wider pool of valuable data: passwords, backup codes, account numbers, PIN hints, card details, identity information, and even cryptocurrency seed phrases. For many users, the notes app has quietly become an informal vault, and malware authors have clearly noticed.

The campaign also reflects a broader truth about mobile fraud: attackers do not always need to break Android itself. In many cases, they can achieve account takeover by tricking users into installing a dropper, granting dangerous permissions, and allowing the malware to observe or control what happens on the device. Google has repeatedly warned that sideloaded apps and accessibility abuse remain major risks in Android threat activity (Android Developers, Google Security Blog).

Technical details

Based on the initial reporting, Perseus is distributed through dropper apps. A dropper typically looks like a legitimate app or update package, but its real job is to install, decrypt, or fetch the main malicious payload after the victim opens it. This separation helps attackers change payloads quickly and can make static detection harder, since the first-stage app may appear less obviously malicious at first glance (CISA).

Although the public summary does not list every capability, malware in the Cerberus/Phoenix line commonly relies on a familiar set of Android abuse techniques:

Accessibility service abuse: Malware requests accessibility permissions, then uses them to read on-screen text, detect which app is open, capture user input, click buttons, approve prompts, and navigate the interface without the victim fully understanding what is happening. Android’s own security guidance highlights accessibility abuse as a serious risk because it can provide broad visibility into device activity (Android Developers).

Overlay attacks: Banking trojans often display fake login forms over real banking or payment apps. Victims think they are entering credentials into a trusted app, but the data is actually captured by the malware. This method has been common across many Android financial malware families, including Anatsa, Octo, Hydra, and others documented by security vendors in recent years (ThreatFabric).

Notification and SMS interception: If a trojan can read notifications or messages, it may capture one-time passwords and transaction alerts. That reduces the protective value of SMS-based MFA and some push-based authentication flows.

Remote control and DTO support: Device takeover means the malware can do more than steal credentials. It can help operators navigate apps, authorize transfers, suppress warnings, and carry out fraud directly from the victim’s device, making activity look more legitimate to bank fraud systems.

The standout Perseus feature is notes-app monitoring. The exact implementation was not detailed in the summary from The Hacker News, but there are several plausible mechanisms. The most likely is accessibility-based monitoring of foreground activity and text content. If the malware can detect when a notes app is opened and read on-screen text or user input events, it can quietly harvest stored secrets. Clipboard theft is another possibility if users copy data into or out of notes. In some cases, local file access may also be relevant, depending on how a notes app stores data and what permissions the malware has.

This matters because notes apps often contain data that falls outside the usual fraud model. A stolen password can lead to account compromise; a stolen backup code can defeat MFA; a stolen seed phrase can empty a crypto wallet; a stolen identity number can support downstream fraud. By expanding collection beyond banking interfaces, Perseus increases the value of each infected device.

No CVE has been cited in the initial report, and that distinction is important. Perseus appears to be an example of malware-driven compromise rather than exploitation of a named Android vulnerability. In other words, the attack path likely depends on social engineering, sideloading, and permission abuse, not on breaking a patched device through a software flaw.

Impact assessment

The immediate victims are Android users who install malicious APKs, especially those who sideload apps from links, third-party stores, or fake update prompts. Users who rely on mobile banking, payment apps, crypto wallets, or who keep sensitive information in notes apps are at the highest risk. Small-business owners are also exposed because many use personal Android devices for banking, invoicing, and account recovery tasks.

The severity is high for infected users. A successful Perseus infection could lead to stolen credentials, intercepted authentication codes, fraudulent transfers, account lockouts, identity theft, and compromise of recovery materials that would otherwise help a victim regain access. If notes apps contain a mix of personal and financial information, the damage can extend beyond a single account.

Financial institutions are affected indirectly but significantly. Device-takeover malware complicates fraud detection because transactions originate from the victim’s own handset, IP ranges, and normal banking app session. That can make malicious activity look closer to legitimate behavior. Banks may see higher fraud losses, more account recovery cases, and pressure to strengthen app-based risk checks and move customers away from SMS OTPs.

For the Android ecosystem, Perseus is another sign that mobile malware operators continue to professionalize. The use of droppers, inherited codebases, and broader data collection suggests an ecosystem where operators iterate quickly and reuse what works. The notes-app angle also shows that attackers are paying attention to user habits, not just app categories.

How to protect yourself

Avoid sideloading whenever possible. Install apps only from trusted sources, and be skeptical of links delivered by SMS, messaging apps, email, or social media. Many Android banking trojans begin with a fake installer or update prompt.

Review accessibility permissions carefully. If a newly installed app asks for accessibility access and its purpose does not clearly require it, deny the request and remove the app. Accessibility access can give malware broad power over the device (Android Developers).

Do not store sensitive secrets in plain notes apps. Avoid keeping passwords, backup codes, PINs, card data, or seed phrases in unprotected notes. If you need to store sensitive information, use a dedicated password manager or a properly protected encrypted vault. For broader privacy on untrusted networks, a reputable VPN service can reduce some exposure, though it will not stop malware already running on your phone.

Move away from SMS-based codes where possible. Prefer passkeys, hardware security keys, or stronger authenticator methods supported by your bank or service. Malware that can read notifications or messages can often bypass SMS OTPs.

Watch for unusual device behavior. Unexpected accessibility prompts, overlays appearing over banking apps, battery drain, apps asking for notification access without a clear reason, or unexplained transaction alerts should all be treated as warning signs.

Keep Android and apps updated. While Perseus does not appear tied to a specific CVE, security updates still reduce exposure to privilege escalation and other supporting techniques. Enable Google Play Protect and keep mobile security tooling active where available (Google Play Help).

If you suspect infection, act quickly. Disconnect the device from networks, contact your bank, change passwords from a clean device, revoke suspicious sessions, and review account recovery settings. If seed phrases or backup codes were stored on the phone, assume they are compromised and rotate what can be rotated immediately.

Perseus is a reminder that Android banking malware is no longer confined to fake bank logins and stolen SMS codes. By extending collection into notes apps, attackers are going after the informal ways people manage secrets on their phones. That makes user behavior, app permissions, and storage habits just as important as traditional anti-malware defenses.

Share:

// FAQ

What is Perseus Android malware?

Perseus is a newly reported Android banking malware family linked to device takeover and financial fraud. Researchers say it builds on techniques associated with Cerberus and Phoenix and is being distributed through dropper apps.

Why is monitoring notes apps so dangerous?

Many users store passwords, backup codes, PIN hints, card details, or crypto seed phrases in notes apps. If malware can read those notes, attackers can steal data that supports account takeover, fraud, and identity theft.

Does Perseus exploit a specific Android vulnerability?

Based on the initial reporting, no specific CVE has been cited. The threat appears to rely mainly on malicious apps, social engineering, and abuse of Android permissions such as accessibility access.

How can Android users reduce the risk?

Avoid sideloading apps, deny unnecessary accessibility and notification permissions, keep secrets out of plain notes apps, use stronger authentication than SMS where possible, and respond quickly to signs of infection.

// SOURCES

// RELATED

Crypto payments linked to human trafficking jump 85%, Chainalysis warns

Chainalysis says crypto payments linked to human trafficking rose 85%, driven by scam compounds and fraud operations using forced labor.

2 min readMar 21

Malicious GitHub Codespaces commands can trigger remote code execution

Researchers say crafted GitHub Codespaces repos or pull requests can run malicious setup commands and expose code, tokens and secrets.

2 min readMar 21

Pre-disclosure attacks climb as nearly 29% of exploited flaws are hit before public disclosure

VulnCheck says 28.96% of exploited flaws in 2025 were attacked before disclosure, up from 23.6% in 2024.

2 min readMar 21

Microsoft disrupts RedVDS, a subscription service tied to AI-powered phishing and BEC attacks

Microsoft says it disrupted RedVDS, a criminal subscription service linked to AI-assisted phishing, BEC and fraud that cost victims millions.

2 min readMar 21