Cybertrading, Airsoft, and ZCB Bamberg: FakeTrade Reporting and Criminal Investigations

In mid-September 2026, the Airsoft case is back in the headlines. An international investigative project called FakeTrade, initiated by DER SPIEGEL and continued by partners such as the Journal de Montréal, portrays an Israeli software company that allegedly supplied trading platforms with technology for years—platforms on which investors believed they were trading while their money was actually being diverted. Bavaria’s Central Office for Cybercrime (ZCB) in Bamberg has been investigating cases of this kind for years. A former Airsoft managing director was convicted there in June 2026.

The story sounds plausible: software built for fraud; call centers contacting customers; investigators cracking the system. A closer look reveals a more complex picture. Journalistic dramatization, the prosecution’s assessment, and what a court actually found are three different things. Conflating them turns a complicated criminal case into a simple morality tale.

What FakeTrade Reports About Airsoft—and What Does Not Follow

DER SPIEGEL’s investigation into Airsoft and cybertrading describes an interplay between a software company, platform operators, and call centers. The Journal de Montréal adds an international perspective. Both articles belong to the same investigative consortium, European Investigative Collaborations (EIC). Their similarities are no coincidence. The articles rely on statements from the same actors—with interests on one side of the dispute!—and on the same documents, and tell the same story.

The narrative is compelling: Airsoft as an all-in-one tool for fake brokers, hundreds of websites, deposits totaling around EUR 950 million, losses in Germany in the tens of millions, and a business model that profits from customers’ losses. Partner reports name platforms such as Fibonetix and Trade Capital, discuss fabricated profits and functions that displayed withdrawals without executing them, and describe the software company receiving a share of deposits.

All of this may be true in individual cases. But it does not yet answer the questions that matter in criminal law: Exactly what did each function do technically? For which platforms has that been specifically established? Who configured which setting? And, above all, what did each accused person know at what point in time?

Journalists naturally have little interest in legal distinctions of this kind—fundamental questions of law. Instead, they portray infrastructure supposedly created “specifically for fraud” and a justice system now successfully dismantling it. It is a powerful story, one the Bamberg prosecution service also likes to tell. It is simply incomplete.

The Airsoft Judgment: Aiding and Abetting Rather Than Joint Perpetration

In June 2026, Bamberg Regional Court sentenced a former Airsoft managing director to four years in prison for aiding and abetting fraud in four cases. The prosecution had charged him as a joint perpetrator.

Joint perpetration means that the person commits the offense together with others as a perpetrator. Aiding and abetting means assisting another person’s offense without committing it as one’s own. The distinction determines the person’s legal role, often the sentence, and how far allegations against software providers can extend in the first place.

The prosecution did not succeed in establishing its broader allegation on this point. The fact that the lead prosecutor continues to defend that position publicly does not turn it into a judicial finding.

Under Section 160(2) of the German Code of Criminal Procedure, prosecutors must also investigate exculpatory circumstances. Someone who tells only the incriminating side of the story does not discharge that duty merely because the narrative sounds coherent.

Goldbeck also describes these networks as organized crime. That may be accurate for individual call-center structures. It does not replace proof that a particular software developer, support employee, or managing director intentionally assisted a specific crime.

Trading Software, Spreads, and Markups: A Feature Is Not Proof of Fraud

A recurring pattern in the reports runs as follows: The software could adjust prices, assign customers to groups, correct trades, or internally mark withdrawals—therefore it was a tool for fraud. Technically, that sounds plausible. Legally, it is a leap in reasoning.

Comparable functions exist in perfectly ordinary brokerage systems. Match-Trader’s documentation describes group-based settings and separate markups on buy and sell prices. DXtrade documents pricing profiles at the level of individual instruments, customer groups, and accounts. The cTrader documentation explains that spreads—the difference between buy and sell prices—can even become negative under market conditions.

Three terms that the reports often confuse:

• A spread is the difference between the price at which you can buy and the price at which you can sell.
• A markup is an adjustment added by a provider to the underlying market price.
• A negative spread and a negative markup are not the same thing.

None of these settings proves, by itself, that prices were fabricated. That would require knowing the price source, the calculation, the contractual terms, and, above all, the prices customers actually saw on their screens. Publicly available documentation from other providers is only comparative material. It establishes neither how earlier Airsoft versions were configured in individual cases nor whether their use was lawful. It shows only that the same functions also appear in lawful software.

The same applies to trade cancellations. In 2010, the U.S. Securities and Exchange Commission reported on rule-based cancellations of clearly erroneous exchange trades. That does not justify arbitrary interference with customer accounts. It does show, however, that “correction” does not automatically mean “deception.” Function, reason, and actual use must be examined separately.

Unfortunately, to date neither the prosecution nor the Regional Court has considered it necessary to have experts examine the technical and economic background of CFD trading—let alone the press, which is content with sweeping and superficial assertions.

Cybertrading Loss Figures: Deposits, Withdrawals, and Proven Damage

The percentage of losing accounts, the payout ratio, and losses caused by fraud must not be equated.

The FakeTrade reports use impressive figures: around EUR 950 million in platform turnover, tens of millions in losses in Germany, and hundreds of websites. Figures like these create the impression that the scale of the case has already been “proved.”

Deposits, turnover, amounts net of chargebacks, and losses judicially established as resulting from fraud are not the same thing. Findings about individual platforms cannot be extrapolated to all customers of a software provider. Large figures require an identifiable period, a defined set of platforms, a currency, and a calculation method. Without those, the figure remains an assertion with journalistic impact but uncertain evidentiary value.

A second misconception concerns loss rates. In 2018, the European Securities and Markets Authority (ESMA) reported that typically 74–89% of retail investor accounts lost money in the CFD offerings it examined. CFDs—contracts for difference—are bets on price movements, often using leverage. Someone who puts in EUR 10 can lose a multiple of that amount if the price moves against them by one percent. High loss rates are therefore a known product risk. They do not establish what proportion of all deposits was paid out.

ZCB Bamberg Investigations: How Digital Evidence Can Mislead

Bavaria’s Central Office for Cybercrime, part of the Bamberg Public Prosecutor General’s Office, handles major cases involving organized structures, substantial IT requirements, and often many victims in several countries. Since October 2024, many of these cases have also been concentrated in a specialized chamber of Bamberg Regional Court.

Chats, server logs, customer databases, and payment flows play a central role in cybertrading cases. A large volume of data does not, however, replace attribution to a person and a specific offense. An IP address, a user account, or a wallet does not identify a person by itself.

For the defense, completeness, translations, the selection of messages presented, and possible alternative explanations therefore matter. Chat excerpts must be read in context. The technical ability to access a system does not prove who made or ordered a particular intervention.

This is precisely where public reporting often becomes imprecise. It quotes individual messages, internal emails, or contractual language and treats them as the full picture. In criminal proceedings, each piece must be attributed to a person, a time, and an act.

International Evidence and the EU E-Evidence Regulation

Platforms, employees, and data are frequently located in different countries. The E-Evidence Regulation (EU) 2023/1543 has applied since August 18, 2026. Subject to certain conditions, authorities can address production orders directly to covered service providers. The deadline is generally ten days, or eight hours in emergencies.

This is a powerful tool. It is not unrestricted permission to access all foreign banking or business records directly. The conditions for an order, its data scope, and available legal remedies remain open to scrutiny.

Where a German arrest warrant exists alongside extradition proceedings abroad, the two proceedings must be coordinated. Information on defending a German case from abroad and working with local counsel is available on Rudolph Rechtsanwälte’s cybercrime page.

Glossary: Cybertrading, CFDs, White Label, and FakeTrade

Cybertrading here refers to online platforms on which customers were supposed to trade foreign currencies, crypto assets, or contracts for difference. In many cases, investigators believe that trading was merely simulated and that deposits ended up with the operators.

CFDs (contracts for difference) contractually track changes in the value of an underlying asset without the customer buying that asset. Leverage magnifies gains and losses.

White label means that one provider supplies the technology and another uses it under its own brand. That alone says nothing about legality or fraud.

FakeTrade is the name of the investigative project, not an established criminal offense.

These terms describe products and functions. Whether an offering was lawful, what information customers received, and whether specific transactions were fraudulent must be examined separately in each case.

Criminal Defense in ZCB Bamberg Cybertrading Cases

Rudolph Rechtsanwälte in Nuremberg advises and defends individuals accused in German cybercrime proceedings. The starting points are access to the case file, examination of the individual allegation, and preservation of exculpatory records. Where technical issues are disputed, the need for an expert must be assessed. International steps are coordinated with lawyers in the relevant country.

Anyone facing an allegation should have their own personal accusation examined before making a statement—not what has been written about “the Airsoft system.”

FakeTrade and the Bamberg Judgment: Individual Guilt Requires Individual Proof

The FakeTrade investigation has brought a real problem to light: investors have been defrauded of substantial sums. A critical reading does not change that. It changes something else: the standard used to infer individual guilt from software, interviews, and large figures.

DER SPIEGEL and its partners describe an infrastructure of fraud. Goldbeck describes a successful investigation into organized crime and maintains an assessment that the court did not adopt in that form. Based on what is publicly known, Bamberg Regional Court found aiding and abetting in four cases, rather than the joint perpetration sought by the prosecution. The resulting press articles often adopt the harsher interpretation and treat it as established.

That is no minor distinction. It is the difference between a dramatic piece of reporting and criminal proceedings in which every person, every function, and every state of knowledge requires its own evidence. Presenting one as the other makes the case simpler than it is.