When the other side produces a phone extraction report, the questions that matter most are how the data was acquired, whether it can be verified, how the timestamps were interpreted, and what the report leaves out. A polished report from Cellebrite or a similar tool can look complete while reflecting a limited extraction, a narrow filter or an unverified parsing result. The checklist below is what we work through when we review opposing forensic work, and most of it can be raised in discovery before an expert ever opens the data.
1. Get the right files, not just the PDF
A PDF or spreadsheet report is a selection of the data, formatted for reading. It is not the extraction. Ask for the full extraction files, the case or project file, and the examiner's notes. A Cellebrite reader file (UFDR) opened in Cellebrite Reader is better than a PDF because it is searchable, but it is still a report generated from the extraction, not the raw image, and the person who generated it chose what went into it.
2. What type of extraction was performed?
Extraction type sets the ceiling on what the report could contain. Ask which method was used and why.
- Logical. Data the device or its backup service makes available, such as messages, contacts and call logs. Limited access to app databases and deleted records.
- Full file system. The device's file system, including app databases, system logs and many artifacts a logical extraction misses. Deleted records within databases are often reachable at this level.
- Physical. A bit level copy of the storage. On modern encrypted phones it is often unavailable or of limited use, so a report claiming a physical extraction of a current device deserves a follow up question.
If a report says no deleted messages were found, the extraction type tells you whether that statement means much.
3. Which tool and which version?
Forensic tools change constantly, and parsers for messaging apps are updated as the apps change. The report should identify the acquisition tool, the analysis tool and the exact version of each. Ask whether the version in use supported the device model, operating system and app versions at issue. NIST's Computer Forensics Tool Testing program publishes test results for mobile acquisition tools, which can be a useful reference point when reliability is challenged.
4. Hash values
A hash value is a digital fingerprint of the extraction files. The report should list hash values generated at acquisition, and those values should match when you verify the files you received. If no hashes were recorded, or the hashes do not match, you have a real question about whether the data you received is the data that was acquired.
5. Chain of custody
- Who seized or received the device, when, and in what condition?
- Was it isolated from networks to prevent remote wiping or incoming data?
- Was it powered on, unlocked or used before acquisition, and by whom?
- Who had access to the extraction files afterward, and where were they stored?
Gaps are not automatically fatal, but they should be explained.
6. Time zones and timestamps
Timestamps are a frequent source of error. Many phone databases store times in UTC, and the report applies an offset for display. Check which time zone the report used, whether daylight saving time was handled correctly, and whether the examiner distinguished between sent, delivered, read and modified times. If the device clock was set manually, or the phone traveled across time zones, the times on the page may not mean what they appear to mean.
7. Parsed, deleted and carved artifacts
Reports mix data recovered in different ways, and the difference affects how much weight each item can bear.
- Parsed artifacts come from active records in a known database structure. They are generally the most reliable.
- Deleted artifacts are records the tool identified as deleted within a database. They may be incomplete, and fields can be misattributed.
- Carved artifacts are reconstructed from raw data without the surrounding structure. Dates, senders and thread assignments for carved items need independent verification.
Ask how each important item was recovered, and whether the examiner checked it against the source database.
8. Filters and exclusions
Reports are usually filtered: by date range, by contact, by keyword, by selected chats or by items an examiner tagged. Filtering is normal, but the report should say what filters were applied. Ask what was excluded and why, whether messages from other apps were considered, and whether context around the selected messages was left out. A conversation looks different when you can see the messages on either side of the excerpt.
9. Validation
Forensic tools interpret data, and interpretation can be wrong. Ask whether the examiner validated key findings by reviewing the underlying database records, comparing results across a second tool, or testing the behavior on a known device. An examiner who cannot explain how a critical artifact was verified is relying on the tool's output alone.
Quick checklist
- Full extraction files, case file and notes requested.
- Extraction type identified and explained.
- Tool names and versions documented.
- Hash values recorded and verified.
- Chain of custody complete.
- Time zone and timestamp handling confirmed.
- Parsed, deleted and carved items distinguished.
- Filters and exclusions disclosed.
- Key findings validated against source data.
We review opposing extraction reports for attorneys, verify the underlying data, run our own analysis where the files allow it, and prepare declarations and testimony on what the report does and does not support. Our methodology page describes how we document our own work to the same standard.
Related services
More articles
- Cellebrite, Magnet AXIOM and Verakey: Why We Use More Than One Forensic Tool
- Validating Text Messages for Court with Digital Forensics
- Authenticating Text Messages in California: What Courts Look For
This article is general information, not legal advice. Every matter is different; contact the examiner to discuss yours.
