On October 4, 2024, the European Court of Justice handed down its Grand Chamber ruling in Bezirkshauptmannschaft Landeck (C-548/21).
The case seems straightforward: Austrian police seized a man’s phone during a cannabis investigation, tried to unlock it without authorization from a prosecutor or court, didn’t document the attempt, and didn’t tell the owner.
And the Court said you can’t do that.
Even attempting to access the data constitutes processing under Directive 2016/680. You need prior authorization from a court or independent authority. You need to document what you did. You need to tell the subject afterward.
These are not suggestions.
As an old investigator, and an early adopter of acquisition tools, I find the results of this ruling very interesting.
It has been called “groundbreaking for investigative work and data protection throughout the European Union.“ It is. You can tell because forensic extraction vendors noticed immediately, in a weird way. Rather than help investigators comply, I see them pouncing on a compliance gap as a market opportunity.
Sigh.
Within months, vendor pitches started arriving on my desk. They want me to think the Landeck ruling means full-device imaging is legally indefensible. Ok, but does that really mean old workflows are non-compliant? We always should have focused on selective extraction tools and targeted, proportionate acquisition. So, asking are your current workflows compliant is a bit like asking how far from spirit you were by torturing the written law.
Investigators were being technically accurate to the letter, while also knowingly dishonest.
The Landeck ruling doesn’t mandate a specific technical approach to data extraction. It mandates a legal process. Prior judicial authorization. Proportionality assessment. Documentation. Notification to the data subject. That’s the stuff of good work.
A department that obtains proper authorization, documents the scope and justification, and notifies the subject afterward can perform a full-device image and remain fully compliant. The legal requirement is the process around the access, not a magic volume of data touched.
Do no harm doesn’t mean tie people up with a garden hose and beat them with a phone book, even if it doesn’t leave physical evidence of torture.
The selective extraction pitch I’m seeing actually inverts the ruling’s logic. I’m not a lawyer but that seems like something lawyers should be looking at sooner rather than later.
The vendor argument goes like this: our tool can target when it pulls, which means the extraction has inherent proportionality, which means skipping heavy authorization processes for a full image. The proportionality is point and click first, ask questions later.
The compliance gap is sold as compliance.
Ugh.
The Court was explicit that even unsuccessful attempts to unlock a phone require prior authorization. I’ve seen that before. Even attempts at data access are sometimes still classified a data breach (e.g. healthcare). The threshold isn’t the volume of data extracted when it’s the act of accessing the device at all. A tool that makes it easier to do quick, targeted pulls at the local station level is a tool that makes it easier to skip the required authorization step. The investigator does a selective extraction, writes up their own proportionality justification after the fact, and calls it compliant because the tool only grabbed what was “relevant.”
That’s the investigator escalating themselves to judge and jury, performing their own prior review, which is exactly what the prior review requirement exists to prevent.
EDRi flagged this dynamic before the ruling even landed. Companies like Cellebrite, MSAB, and Grayshift sell extraction equipment widely deployed at corporations and police stations for prosecuting any and all crimes, including petty ones. I still remember where my classmates came from for EnCase certification.
The bottom line is that technology outpaced the legal frameworks years ago. In many EU countries, smartphones still end up treated as ordinary evidence-gathering, another object to be seized, without special data considerations.
Germany is the clearest example. The rules on search and seizure in the Code of Criminal Procedure (StPO) don’t differentiate between complex digital data carriers and other objects. It’s left entirely to the interpretation practice of prosecutors and investigating judges to apply proportionality case by case. In practice, suspects’ laptops and smartphones are often seized even when suspicions are tenuous and the hope of finding evidence is based solely on investigative experience.
Landeck is supposed to fix this.
The Court laid down clear requirements: national legislatures must define with sufficient precision the factors to be taken into account, including the nature and categories of offences. Access must be subject to prior review by a court or independent authority except in urgent cases. The data subject must be informed of the grounds for authorization as soon as disclosure won’t compromise the investigation.
Yet the vendor ecosystem is openly flogging workarounds in real time. Selective extraction tools to bypass Landeck, marketed as Landeck-compliant by design, allow investigators to access device data faster and with less friction than a full forensic image would require. The reduced scope becomes the justification for reduced oversight. The tool tries to avoid the authority of a judge.
This is a pattern we’ve seen before. Every time a court or regulator establishes a constraint on data access, the compliance industry immediately begins selling products that satisfy the letter of the requirement while undermining its purpose. Cookie consent banners. Privacy impact assessments. And now, “proportionate” mobile forensic tools for unauthorized access to feel authorized because the extraction was selective.
The Landeck ruling said: get a judge involved before you touch the device. It didn’t say buy a tool that makes touching the phone feel proportionate enough that you can ignore Landeck.
Rubber hose cryptography still counts as a threat to confidentiality.
Investigators who want to comply with the actual Landeck ruling should focus on the procedural requirements: get authorization before access, document scope and justification, notify the subject when appropriate. The tool matters less than whether a judge reviewed the request for acquisition.
The ones buying selective extraction tools to skip authorization aren’t achieving compliance. They are being sold a black bag of plausible deniability.