security

Why 'Secure' File Transfer Isn't Enough Without Access Security

ShinyHunters stole millions of student records from a 'secure' file transfer service in 2026. Transport security alone doesn't protect against stolen login credentials.

Between May 22 and September 18, 2026, an attacker used a single compromised login credential to access a secure file transfer service used by the Educational Testing Service (ETS). During those four months, files containing personal information belonging to millions of California students were copied and removed. ShinyHunters — the threat actor group responsible for several major data incidents in recent years — claimed credit.

The service used in this incident was designed for secure file transfer. It used encrypted connections. It was built for organizations that handle sensitive data. The security worked as designed. The attacker did not break the encryption or exploit a technical vulnerability in the transfer protocol. They simply logged in.

This incident is a useful lens for understanding a misconception that affects how many people and organizations think about storing and sharing sensitive personal files: the confusion between transport security and access security.


What “Secure” Means When a File Transfer Service Says It

When a file transfer service describes itself as secure, it is typically describing properties of the data in transit: the connection between you and the service is encrypted, the file is not readable by someone intercepting the network traffic, the data cannot be tampered with during transfer without detection.

These are real and important protections. In 2026, any reputable service handling sensitive files should offer TLS 1.3 for transport security, and the absence of that would be a serious problem.

But transport security protects the channel. It says nothing about who has the key to the front door.

Access security is about authentication: who is allowed in, how they prove it, and what happens when those credentials are stolen. The ETS breach was an access security failure. A credential — a username and password, or possibly an API key or session token — was stolen or guessed. The attacker used that credential to authenticate as an authorized user. From the service’s perspective, the access looked legitimate. The transport was encrypted. The file copies left no obvious anomaly in a simple log review.

The files were simply gone by the time anyone noticed.


The Long Window Problem

What makes the ETS incident particularly instructive is its duration. The unauthorized access began on May 22 and was not identified until after September 18 — approximately four months.

A four-month window means that:

  • The attacker had time to access files selectively and carefully, which reduces the chances of detection via volume-based anomaly detection
  • The breach was identified not through internal monitoring but through claims by the attacker group
  • There is no way to know precisely which files were accessed in any given week during that period
  • Even if you change your password today, files that were accessible during the window were already exposed

This is not unusual in credential-based breaches. When an attacker has valid credentials, they can access files at any time without triggering the alerts that a brute-force or injection attack might. If the service does not have behavioral monitoring — looking at whether this user is accessing files at unusual times, downloading unusual quantities, or accessing files they have never touched before — the intrusion can persist indefinitely.

Research consistently shows that the average time between a breach and its detection is measured in months. For credential-based breaches specifically, the figure is often higher, because the attacker generates no technical anomalies — only behavioral ones, which require more sophisticated monitoring to catch.


Why One Compromised Credential Is Enough

The natural question is: why did a single compromised credential have access to files belonging to millions of students?

The answer is that many file storage and transfer systems grant broad, account-wide access rather than object-level or purpose-specific access. An account authorized to access the ETS file transfer service was authorized to access the files in it — not just specific files, not just files from a specific time period, not just files related to a specific project.

This broad-access model is common in enterprise systems. It is also common in consumer cloud storage: an account credential for Google Drive gives access to everything in that Drive, not just specific files. A compromised account credential is therefore a key to all of its contents.

More granular access control — the principle of least privilege — is the design alternative. Under this model, a credential grants access only to the specific resources it needs: this folder, not the entire storage; this project’s files, not the entire account. A compromised credential in this model exposes only what that specific credential could reach.

Implementing least-privilege access is straightforward in principle and consistently underimplemented in practice. It requires active design decisions about what each access grant needs, and it creates friction compared to “grant full access to everything.” Most systems default toward broader access.


What Secure File Transfer Does and Doesn’t Protect

To use these properties correctly when evaluating services, it helps to be clear about what each protection actually covers.

Transport encryption (TLS): Protects against network interception. An attacker positioned on the same network — a shared Wi-Fi network, a compromised router, an internet service provider — cannot read your files while they are in transit. This is important protection. It does not protect against anything that happens after the file reaches the server.

Encryption at rest: Protects against an attacker who gains access to physical storage media. The files on the server’s disks are encrypted and not readable without the keys. It does not protect against anyone who can authenticate to the service, including the service provider itself.

Password-protected links: Protects against someone who finds a link to your file without knowing its password. It does not protect against the service’s own infrastructure, legal requests to the service, or an attacker who has your account credentials.

File expiration: Limits the window during which a file can be accessed if a link or credential is compromised. If a shared file expires after 48 hours, a compromised credential cannot access it a month later. This is underused protection that narrows the window of exposure from a delayed-detection breach.

Multi-factor authentication: Protects against password compromise specifically. An attacker who steals only a password cannot log in without also having access to the second factor. This is one of the most reliable defenses against credential-based breaches. It does not protect against more sophisticated attacks on the second-factor mechanism itself, but it eliminates the largest category of credential-based access.


Evaluating a Service for Sensitive Personal Files

If you store or share sensitive personal files — medical records, legal documents, identity documents, financial records, personal communications — the security properties you need go beyond “this service uses encryption.”

The questions worth asking:

Does the service require multi-factor authentication? Not just offer it as an option, but require it. A service where MFA is optional and most users have not enabled it provides substantially weaker protection than one where it is mandatory.

What does the service’s login monitoring look like? Are you notified of logins from new devices or unusual locations? Can you see a history of account access? Visibility into your own account activity is the first line of defense against credential-based access you did not authorize.

How does access scope work? If someone gains access to your account, how much of your stored data can they reach? A service that allows folder-level access grants, where specific collaborators can only see what they need to see, is structurally more resistant to the broad-access problem.

What is the breach history and detection capability? Has the service disclosed past incidents, and how long was the window between the breach and detection? A service that detected an intrusion within hours is operating very differently from one that learned about it months later from the attacker.

Does the service log access at the file level? Knowing which files were accessed, when, and from where — and being able to audit that history yourself — is valuable both for detecting unauthorized access and for understanding the scope if a breach does occur.


The Personal File Storage Angle

The ETS breach involved an institutional system and affected third parties — the students whose records were stored there. But the patterns it illustrates apply directly to personal cloud storage.

Most people store sensitive files in cloud services and rely entirely on the service’s security architecture, without actively monitoring their own access logs, enabling available security features, or reviewing what they have stored and who has access.

A compromised credential for Google Drive or iCloud exposes not just recent files but everything stored in those accounts — years of documents, photos, and records. The credential may be compromised through a phishing attack, a breach of another service where the same password was reused, or a device that has been compromised.

The practical steps that make the biggest difference:

Enable multi-factor authentication on every cloud service you use. This single step eliminates the most common path to credential-based unauthorized access.

Use a password manager and use a unique password for every service. Password reuse means a breach at one service can unlock another. A unique password for your cloud storage account means that its exposure requires specifically targeting that service, rather than being found in a credential database from an unrelated breach.

Review your account sign-in history periodically. Most services maintain a history of logins with device type and location. Unfamiliar entries are worth investigating.

Store sensitive files in services with appropriate security for what they contain. A file transfer service optimized for speed and convenience is a different security environment than a purpose-built storage service with stronger access controls.


The Correct Mental Model

The takeaway from the ETS breach is not that file transfer services are untrustworthy. It is that transport security and access security are different things, and the “secure” label in a service’s name or marketing describes only one of them.

When you store a sensitive file somewhere, ask: if my password for this service were compromised tomorrow, what could an attacker access, for how long, without my knowing? The answer to that question tells you more about your practical exposure than whether the connection is encrypted.

daftei secures files with TLS 1.3 in transit and AES-256 at rest, and provides standard account security including activity monitoring. It is designed for intentional personal file storage — not bulk file transfer — with GDPR and CCPA compliance, no data selling, and a 30-day grace window on account deletion before permanent, irreversible erasure. Using a service you have actively chosen for its security posture, with MFA enabled and a unique password, is a different situation than using whatever was convenient when you needed to send a file.

The difference is deliberate access security, not just encrypted transport.

Your memories deserve better than an ad platform.

Try daftei free →
← All posts