
Dark web research, OSINT and eDiscovery in digital forensics
Introduction: when the trail leads into hidden places
Digital investigations rarely end at the surface of the visible internet. In a growing number of cyber attacks, white-collar crime cases and internal corporate investigations, the decisive traces lead deep into areas that cannot be reached with conventional search engines and investigative methods: the dark web, decentralised communication channels, encrypted marketplaces and the sprawling network of publicly accessible yet hidden information sources.
For forensic experts, law enforcement, law firms and compliance departments, three disciplines have become indispensable: dark web research, OSINT (open source intelligence) and eDiscovery. Each follows its own technical and legal rules – together, they form a powerful toolkit for investigating digital crime and securing evidence that holds up in court.
This article explains how these methods work, where their legal limits lie and why applying them correctly decides whether an investigation succeeds or fails.
Dark web research
What the dark web is – and what it is not
In public perception, the dark web is often confused with the so-called “deep web”. The distinction matters both forensically and legally:
- Surface web: the part of the internet indexed by search engines – only a small part of the content available online.
- Deep web: non-indexed content such as database queries, password-protected areas, internal corporate networks and medical records. Not an illegal space as such.
- Dark web: a subset of the deep web that can only be reached with special software such as Tor (The Onion Router), I2P (Invisible Internet Project) or Freenet. Deliberately anonymised, technically demanding to access, and home to both legitimate activity (press freedom, whistleblowing) and criminal structures.
The forensic relevance of the dark web stems from its role as the primary marketplace for:
- Stolen corporate data and credentials (credential dumps)
- Ransomware-as-a-service (RaaS) platforms
- Leak sites of ransomware groups (double extortion)
- Stolen credit card data and identity information
- Cyber weapons, exploits and zero-day vulnerabilities
- Services for hire: DDoS, espionage, data exfiltration
Technical foundations of dark web research
The Tor network and .onion services
The Tor network routes traffic through several encrypted relay nodes (entry guard, middle relay, exit node) so that the original IP address is concealed. So-called hidden services (.onion addresses) make it possible to run servers whose physical location also remains hidden.
Forensically, this means that direct IP tracing is practically impossible without the cooperation of relay operators or authorities. Dark web research therefore focuses on:
- Metadata analysis: timestamps, writing styles, mistakes in operational security (OpSec)
- Crypto wallet tracing: Bitcoin and other cryptocurrencies leave traces on the blockchain
- Correlation analysis: matching activity patterns in the Tor network with clearnet activity
- Anonymisation failures: many criminals make classic OpSec mistakes – wrong time zone settings, reused usernames, identical writing styles
I2P and alternative networks
I2P (Invisible Internet Project) differs from Tor in its fully decentralised, packet-based routing without exit nodes. It is designed primarily for internal communication and services. Forensically, it is harder to access than Tor, but also less widespread.
Freenet, by contrast, is a censorship-resistant peer-to-peer network for anonymous file sharing – forensically relevant in cases of illegal content sharing.
Methods of forensic dark web research
Passive research (legal and safe)
Passive dark web research covers the observation and documentation of publicly accessible dark web content without any active intervention:
- Monitoring leak sites for stolen corporate data
- Watching dark web forums for mentions of the target organisation
- Analysing credential dumps for affected email domains
- Documenting ransomware negotiation sites
- Screenshot- and hash-based preservation of evidence from .onion sites
Active research – legal limits
Active research – deliberately infiltrating dark web communities, creating fake accounts or buying illegal goods for evidence purposes – sits in a legal grey area and is generally not permitted in Austria without official authorisation.
The limits are clear: the provisions of the Austrian Criminal Code (StGB) apply to forensic investigators too – for example Section 118a (unlawful access to a computer system), Section 126c (misuse of computer programs or access data) and Section 148a (fraudulent misuse of data processing). What counts is exclusively lawful, reproducible and court-admissible preservation of evidence.
Cryptocurrency tracing
Cryptocurrencies are pseudonymous, not anonymous. Every transaction is permanently recorded on the blockchain. Forensic blockchain analysis makes it possible to carry out:
- Clustering: grouping wallets that belong to the same person
- Taint analysis: following funds through mixers and tumblers
- Exchange identification: detecting transactions to regulated exchanges with KYC obligations
- On-chain attribution: linking wallet addresses to known actors
Tools such as Chainalysis, Elliptic and CipherTrace are used by authorities and private forensic teams worldwide.
Dark web research in practice
Ransomware incidents: after an attack, we check whether the organisation is already listed on a leak site, which data has been published and whether negotiation channels are active.
Credential monitoring: regular monitoring for stolen credentials makes it possible to identify compromised accounts before attackers use them.
Threat actor attribution: by analysing communication patterns, technical artefacts and OpSec mistakes, attackers can be attributed with high probability to particular groups or individuals.
Due diligence: in M&A processes, we check whether target companies have already been the subject of data leaks or dark web activity.
OSINT – open source intelligence
Definition and scope
OSINT refers to the systematic collection, processing and analysis of information from publicly accessible sources. The term originated in a military and intelligence context, but has developed a meaning of its own in digital forensics, corporate investigations and cyber security.
The key point: OSINT is legal. Only information that is accessible without circumventing technical or legal safeguards is used. The skill lies not in gaining access, but in systematically connecting and analysing this information.
OSINT sources at a glance
Social networks and digital identities
- LinkedIn, XING, Twitter/X, Facebook, Instagram
- GitHub (source code leaks, comments, commit histories)
- Reddit, specialist forums
- Employer review sites (Glassdoor, Kununu)
Forensic relevance: employees often unknowingly publish sensitive information – organisational charts, technology stacks, internal processes, personal data.
Public registers and databases
- Firmenbuch (Austrian companies register), Handelsregister (Germany), Companies House (UK)
- Land registers and property registers
- Insolvency registers
- Databases of court decisions
- WHOIS databases for domain information
- Certificate Transparency logs (crt.sh) – every SSL certificate ever issued
- Shodan, Censys, ZoomEye – search engines for internet-connected devices
Technical OSINT sources
- DNS history: previous IP addresses, server structures
- Wayback Machine / archive.org: historical versions of websites
- Pastebin and similar services: frequently the first place data leaks appear
- Job postings: a company’s technology stack and security strategy
- Document metadata: author names, software versions, GPS coordinates in images (EXIF)
Geolocation and image analysis
By analysing image backgrounds, shadows, reflections, street signs and other visual clues, OSINT techniques allow people or events to be located geographically with precision – a method used by organisations such as Bellingcat to uncover facts in international conflicts.
OSINT frameworks and tools
Maltego: visualises connections between people, domains, IPs and organisations. The industry standard for complex OSINT analysis.
Recon-ng: a modular reconnaissance framework for automated OSINT collection.
theHarvester: email addresses, host names and subdomains from public sources.
SpiderFoot: an automated OSINT platform with more than 200 modules.
OSINT Framework (osintframework.com): a structured overview of all relevant OSINT sources and tools.
Shodan: the “search engine for the Internet of Things” – it finds exposed cameras, routers, industrial control systems and database servers.
OSINT in forensic practice
Researching individuals
For forensic purposes – such as identifying unknown attackers or vetting suspects – OSINT combines:
- Digital footprints (usernames, email addresses, phone numbers)
- Social connections and networks
- Historical online activity
- Technical artefacts (browsers used, timestamps, language patterns)
Corporate investigations
In internal investigations – suspected data theft, corruption, breaches of competition rules – OSINT provides context:
- Links between employees and competitors
- Unusual activity on social networks
- Financially conspicuous lifestyle indicators
Threat intelligence
OSINT is the foundation of modern threat intelligence:
- Identifying attacker infrastructure (C2 servers, phishing domains)
- Early warning of planned attacks by monitoring hacker forums
- Attributing attacks to known threat groups
Legal limits of OSINT in Austria
OSINT operates in the tension between freedom of information and data protection law (GDPR and the Austrian Data Protection Act, DSG). Key points:
- Scraping public data is generally permitted as long as no technical safeguards are circumvented.
- Further processing of personal data is subject to GDPR principles – even if the data was publicly accessible.
- For evidence purposes, OSINT results must be reproducible and documented – screenshots alone are often not enough in court. Hash values, timestamps and archiving logs are required.
- Sections 119 et seq. of the Austrian Criminal Code (violation of the secrecy of telecommunications) set clear limits on monitoring communications.
eDiscovery – electronic evidence discovery
What is eDiscovery?
eDiscovery (electronic discovery) is the process of identifying, collecting, preparing, reviewing and producing electronic data as evidence in legal proceedings. The term comes from the US legal system, but with the digitalisation of legal practice it has become central in Austria, Germany and across the EU as well.
eDiscovery is not a single tool but a structured process – known as the EDRM (Electronic Discovery Reference Model):
- Information governance – proactive management of electronic information
- Identification – determining relevant data sources
- Preservation – legally sound preservation to prevent spoliation
- Collection – forensically correct collection
- Processing – preparation, de-duplication, indexing
- Review – reviewing for relevance and privilege
- Analysis – pattern recognition, timelines, entity extraction
- Production – handover in legally accepted formats
- Presentation – presentation in court
Data sources in the eDiscovery process
Modern eDiscovery has to cover an ever-growing range of data sources:
Traditional sources:
- Email (Exchange, Outlook, Gmail, Lotus Notes)
- Documents (Word, Excel, PDF, CAD files)
- Databases
Modern sources:
- Microsoft 365 / Teams: chats, channels, files, calendar entries
- Slack, WhatsApp Business, Signal: business communication on messaging platforms
- SharePoint, OneDrive, Google Drive: cloud storage
- CRM systems: Salesforce, HubSpot
- ERP systems: SAP, Oracle
- Mobile devices: SMS, call logs, app data
- Social media: corporate accounts, employee activity
- IoT devices: access control systems, cameras, sensor data
Technical challenges of eDiscovery
Data volume and variety
Organisations produce terabytes of data every day in dozens of formats. Without specialised tools, manual review is impossible. Modern eDiscovery platforms use:
- Predictive coding / technology-assisted review (TAR): machine learning for automated relevance assessment
- Near-duplicate detection: identifying almost identical documents
- Email threading: reconstructing complete email conversations
- Entity extraction: automatic recognition of people, places, dates and financial figures
Metadata and its forensic significance
In eDiscovery proceedings, metadata is often more decisive than the content of a document itself:
- Creation date vs. modification date: reveals manipulation
- Author metadata: who really created a document?
- Print data: when was a document printed?
- Version history: all earlier versions of a document
- GPS coordinates: in photos and certain document types
Encryption and password protection
Password-protected documents and encrypted communication pose considerable challenges for eDiscovery. Specialised forensic tools can crack passwords or bypass encryption in many cases – legally, however, only under specific conditions and with the appropriate authorisation.
eDiscovery in the Austrian legal context
Austria has no comprehensive disclosure obligation comparable to the US discovery system. Nevertheless, eDiscovery is relevant in several contexts:
Civil proceedings
In Austrian civil procedure (ZPO), the court can order the production of documents (sections 303 et seq. ZPO). Electronic documents count as documents. Refusing to produce them can lead to disadvantages in the burden of proof.
Employment law investigations
In internal investigations into suspected employee misconduct (data theft, corruption, breaches of confidentiality), eDiscovery measures must be reconciled with data protection law and the Austrian Labour Constitution Act (ArbVG). A works agreement or consultation with the works council is required in many cases.
Regulatory requirements
- NIS2 / NISG 2026: reporting obligations for security incidents require fast data collection and analysis.
- GDPR Art. 33/34: notification obligations for data breaches within 72 hours.
- Financial market regulation (FMA): retention obligations for communications.
International proceedings
In proceedings with US counterparties or before US courts, Austrian companies are effectively subject to the US discovery regime. Transferring personal data to the US is governed by the CJEU’s Schrems II ruling and requires appropriate safeguards (standard contractual clauses, BCRs).
eDiscovery platforms in professional use
Relativity: the industry standard for large eDiscovery cases. Cloud-based, AI-supported, scalable.
Nuix: particularly strong at processing large, heterogeneous data sets. Used in forensic investigations worldwide.
Exterro: an integrated legal GRC platform with a strong focus on compliance.
Microsoft Purview (formerly Advanced eDiscovery): native integration with Microsoft 365 – an efficient solution for organisations that rely primarily on Microsoft products.
Magnet AXIOM: primarily forensics-oriented, but with strong eDiscovery features for mobile devices and cloud sources.
How we use these methods in client cases is described on our eDiscovery page.
Working together – dark web, OSINT and eDiscovery in complex investigations
The three disciplines can rarely be used in isolation. In practice they interlock – and their effectiveness comes precisely from the combination.
Case example: ransomware attack with data exfiltration
Phase 1 – immediate response and eDiscovery: after a ransomware attack, the first task is to determine the extent of data exfiltration. eDiscovery techniques help identify the affected data systems, evaluate logs and reconstruct the attack path.
Phase 2 – dark web research: in parallel, the dark web is monitored for the attacker group’s leak sites. Has data already been published? Is there a negotiation site? Which data was stolen?
Phase 3 – OSINT attribution: by analysing the malware signatures used, the infrastructure (C2 servers, registrar, hosting) and communication patterns, the aim is to identify the attacker group and match it against known threat actors.
Phase 4 – court-admissible documentation: all findings are secured according to forensic standards, hashed and summarised in an expert report that meets the requirements of Austrian law enforcement and courts.
Case example: internal investigation – suspected data theft
A company suspects that a departing employee has stolen customer data and passed it on to a competitor.
OSINT: LinkedIn analysis shows a new position at the competitor. The employee’s public GitHub commits contain code similar to internal projects.
eDiscovery: analysis of email logs, USB activity, cloud sync logs and access logs for sensitive files in the weeks before the resignation.
Dark web research: checking whether internal data has already appeared in credential dumps or on data marketplaces.
Result: a complete, court-admissible picture of the offence – or the exoneration of the suspect.
Legal and ethical framework
Chain of custody
Every forensic measure in dark web research, OSINT and eDiscovery is worthless if the chain of custody – the complete traceability of every action taken on evidence – is not maintained.
In concrete terms, this means:
- Documenting every access to evidence with a timestamp
- Hash values before and after every processing step
- Using forensically valid, write-protected copies
- Four-eyes principle for critical actions
GDPR and data protection
Investigations involving personal data – which is practically every investigation – must be GDPR-compliant. That means:
- Proportionality: only the data necessary for the purpose of the investigation may be collected.
- Purpose limitation: data may not be used for other purposes.
- Deletion obligations: once the investigation is complete, personal data must be deleted unless there is a legal retention obligation.
- Data subject rights: under certain conditions, data subjects must be informed about the processing – unless this would jeopardise the purpose of the investigation.
Admissibility in court
In Austrian criminal proceedings, the court decides on the admissibility of evidence under section 166 of the Code of Criminal Procedure (StPO). Unlike in US law, unlawfully obtained evidence can, under certain circumstances, still be used, but is subject to a balancing test. In civil proceedings the situation is more restrictive.
For forensic practice, the rule is: lawful methods, careful documentation and reproducible results are the only path to court-admissible evidence.
Conclusion: complex methods require certified expertise
Dark web research, OSINT and eDiscovery are not tasks that can be handled with a browser and some basic technical knowledge. They require:
- In-depth technical knowledge of networks, protocols, data formats and forensic tools
- Legal expertise in Austrian and European data protection and criminal law
- Experience in dealing with authorities, courts and insurers
- Certifications such as CISSP, SANS GCFE, GCFA or specialised eDiscovery certificates
- Court-admissible documentation in line with ISO/IEC 27037
A single mistake in preserving evidence can mean that months of work fail in court – or that an offender walks free because the evidence is inadmissible.
DFIR Austria covers the full range of these disciplines. As a court-certified and sworn expert witness, Michael Meixner leads investigations that are not only technically precise but also documented in a legally sound way – see digital forensics.
Do you have questions about a specific situation? Send us your enquiry now – confidential, without obligation and under NDA on request.



