Did Iran Take Its Cue from China? The Barati Arrest and the Mabna Institute

"One of the largest state-sponsored hacking campaigns ever prosecuted." — US Department of Justice on the 2018 indictment of nine Iranian nationals, quoted in Kim Zetter, Zero Day, 1 July 2026

One detail in Kim Zetter's report deserves the attention of every research CIO in Europe. The 2018 US indictment of nine Iranian nationals names 144 universities in the United States and a further 176 across nearly two dozen other countries — Germany explicitly among them, alongside Australia, the UK, China, Israel and Japan. Individual institutions are not identified in the public record. But given the target disciplines — science and engineering, technology, social sciences, medicine — German institutions being affected is a question of probability rather than possibility.

The occasion for the report is the arrest of a 39-year-old Iranian-Turkish national in Montenegro last week at the request of the FBI. Turkish media identify him as Amir Barati; Montenegrin police released only initials. He is accused of attacking more than 150 US universities since 2013 to steal data on behalf of Iran's Islamic Revolutionary Guard Corps (IRGC) and Iranian universities. The damages cited come to 3.4 billion US dollars, though the police statements do not explain how that figure was reached.

An important caveat: Barati is not named in the 2018 indictment, and no public charging document against him exists in the New York court where he is reportedly to be prosecuted. Zetter is careful — his case may be connected to the 2018 proceedings, and the damages figure matches the one cited there against the Mabna Institute in Tehran. An extradition hearing was scheduled in Montenegro this week.

Zetter's thesis, and why it matters in Europe

Zetter draws a parallel worth taking seriously. Prosecutors say the Mabna Institute was founded "in approximately 2013", for the express purpose of stealing academic data and intellectual property abroad on behalf of Iranian government bodies, universities and private clients. That same year, Mandiant published the APT1 report, the first comprehensive public exposure of Chinese economic espionage.

Zetter puts the conclusion explicitly as an open question: did Tehran take its cue from the Chinese example? She does not answer it. The timing is striking; the record supports no more than that.

Until then, visible Iranian cyber activity abroad had a different character: DDoS campaigns against US banks, the intrusion into a small dam in New York State, ransomware against 200 victims, a disinformation campaign aimed at US voters in the 2020 election, intrusions against Trump campaign staff. Systematic theft of intellectual property focused on academic research has not traditionally been an Iranian hallmark.

That is exactly what shifts the European threat picture. German domestic intelligence has warned for years about Chinese innovation extraction and engages universities directly through its economic-security programme. The Iranian component barely features — although it targets the same sector that carries German innovation: Fraunhofer, Helmholtz, Max Planck, DFG-funded consortia, technical universities.

The attack path is unremarkable, which is why it works

Nothing in the 2018 indictment's account of the vector is surprising. That is not reassuring; it is the opposite. The attackers sent spear-phishing mail appearing to come from colleagues at other universities. The messages praised recipients for a recently published paper and carried a link to a prepared page that harvested their university credentials.

Of more than 100,000 accounts targeted, nearly 8,000 were compromised, roughly half belonging to professors in the US. Through those accounts the attackers reached theses, dissertations, research data and other intellectual property. On compromised accounts they also set up automatic forwarding, so that incoming and outgoing mail was read continuously.

The attack needs no zero-days. It needs patience, careful target research, serviceable English, and enough targets for the hit rate to add up.

The genuinely instructive part: the business model

Universities were not the only targets. The indictment adds 30 US companies, five federal and state agencies — the Department of Labor, the Federal Energy Regulatory Commission, the states of Hawaii and Indiana — and the United Nations. For those victims the objective was mostly wholesale theft of mailboxes. Among them were three academic publishers, eleven technology firms, five consultancies, a law firm, two banking or investment firms, a healthcare company and a biotech company.

The most revealing detail sits at the edge of the indictment: access to hijacked academic accounts was passed to the IRGC — or sold to Iranian universities, which used it to reach library resources and subscription databases.

That is not an intelligence programme in the classical sense. It is a service business with a state as its principal customer. In total, prosecutors say, the institute stole more than 31 terabytes of data. The distinction matters defensively: a contractor paid by volume works broadly and takes every reachable target along the way.

Iran is not China — what that qualifies, and what it does not

Zetter closes on an important difference. Chinese operations run on highly skilled operators, a constellation of shell companies and a clear hierarchy tied to the People's Liberation Army or the Ministry of State Security. Iranian efforts look less mature by comparison.

Lower maturity is not lower impact. More than 31 terabytes out of a single contracting firm is a plain statement. The maturity question still matters for defensive planning: an institution equipped to counter Chinese operations will stop Iranian ones at a fraction of the effort. But lower maturity is not lower ambition, and the ambition is strategic — technological catch-up under sanctions.

What research institutions should do

Four measures follow directly from the 2018 indictment:

1. Harden library and publisher logins. Shibboleth SSO with hardware-backed MFA is the only robust answer to credential phishing. Reaching JSTOR, Elsevier or Springer with a username and password in 2026 means losing to an attack path unchanged since 2013 — all the more so given that access to exactly these resources was what Mabna sold.
2. Monitor forwarding rules actively. In Exchange and Google Workspace, every newly created rule should raise an alert, and external forwarding belongs behind a policy block rather than a recommendation.
3. Aim awareness training where the access sits. The targets are not vice-chancellors but postdocs, doctoral candidates and research staff. And the lure is not a parcel notification; it is praise for their own publication.
4. Bring in the national agencies. In Germany, both the BSI and the Verfassungsschutz economic-security programme advise at no cost. A phone call is cheaper than a terabyte of lost raw data.

What the Barati case additionally shows

Arrests in these cases are the exception, because Iran and the United States have no extradition treaty — suspects can only be detained once they leave the country. The 2018 docket consists largely of sealed filings, and whether anyone has been arrested since the indictment is not publicly known. Zetter notes that further sealed documents were added in 2020, 2022 and most recently in December 2025. Those may reflect arrests not yet made public, or simply detail on what was stolen and what it cost.

For defenders this is sobering rather than reassuring. Operations very probably continue while the case files accumulate. The only lever that actually improves an institution's position is its own hardening, not the prospect of Western prosecution.

Read that way, Zetter's piece is a useful mid-2026 reality check: anyone still filing the Iranian threat under DDoS and ransomware is running more than a decade behind.