Skip to main content
Mallory
← Back to News
leaked-secret-api-keycredential-access-methodeducation-sector-threatai-platform-security

Attacker Uses 1,438 Fake FOFA Accounts to Harvest Exposed OpenAI and DeepSeek API Keys

Breach & Incident Intelligence · 2 sources · Updated 7d ago · First seen Nov 10, 2026

Attacker Uses 1,438 Fake FOFA Accounts to Harvest Exposed OpenAI and DeepSeek API Keys

An unidentified threat actor automated the creation of at least 1,438 fake FOFA accounts to locate and harvest exposed OpenAI and DeepSeek API keys from misconfigured internet-facing servers, according to VECERT. The operation extracts credentials, checks their validity and remaining credit, and uses working keys to power the actor’s AI coding assistant and other scripts. Affected organizations face unauthorized AI consumption, depleted credits, and unexpected charges.

VECERT tracked the activity as VECERT-TI-2026-1107-B and linked it to Operation Ghost-Campus, tentatively assessing that the actor may be based in China. The same attacker server reportedly hosted tools that collected sensitive information about 5,132 students at Shanghai Publishing and Printing College and automated fraudulent reading activity to earn school points. Organizations should revoke exposed keys, review usage logs, prevent secrets from being publicly accessible, and set AI spending limits. Reported infrastructure indicators contain an unresolved discrepancy between two purported mail-server IP addresses and should be validated before operational use.

Is this vendor in your chain?See attribution, scope, and your downstream exposure.
Check the blast radius
Event timeline

How this story unfolded

Events from the most recent confirmed update back to the earliest known activity.

6 events
  1. Reported Nov 10, 20267d ago

    VECERT uncovers the toolset and links it to Operation Ghost-Campus

    VECERT analyzed files recovered from the attacker-controlled server and tracked the investigation as VECERT-TI-2026-1107-B, linking the activity to Operation Ghost-Campus. It tentatively assessed that the actor may be based in China and reported technical findings and infrastructure indicators, although two listed mail-server IP addresses conflicted.

    Hackers Use 1,438 Fake FOFA Accounts to Steal OpenAI and DeepSeek API Keys Across the Internet | Cryptika Cybersecurity
  2. Reported Nov 10, 20267d ago

    Associated bots manipulate college reading-activity points

    Tools on the same attacker server reportedly simulated reading activity every 30 seconds to earn school points. The bots accumulated approximately 10 points every 11 minutes.

    Hackers Use 1,438 Fake FOFA Accounts to Steal OpenAI and DeepSeek API Keys Across the Internet | Cryptika Cybersecurity
  3. Reported Nov 10, 20267d ago

    Attacker reportedly collects sensitive information on 5,132 college students

    VECERT reported that the same attacker collected information about 5,132 students at Shanghai Publishing and Printing College's Fengxian campus, including names, phone numbers, national ID numbers, home addresses, and campus card balances. The associated scraper reportedly refreshed the collection every 48 hours.

    Hackers Use 1,438 Fake FOFA Accounts to Steal OpenAI and DeepSeek API Keys Across the Internet | Cryptika Cybersecurity
  4. Reported Nov 10, 20267d ago

    Stolen AI access powers the attacker's coding assistant and scripts

    According to VECERT, the attacker used stolen API access to power Pi — AI Coding Assistant, an interactive web terminal built on GoTTY. Other scripts on the attacker's server also used the stolen access.

    Hackers Use 1,438 Fake FOFA Accounts to Steal OpenAI and DeepSeek API Keys Across the Internet | Cryptika Cybersecurity
  5. Reported Nov 10, 20267d ago

    Automated searches harvest and validate exposed OpenAI and DeepSeek keys

    The operation searched exposed .env files, Docker settings, and poorly secured Git repositories, extracting API keys and checking their usability and balances. Recovered evidence showed 16 verified credentials and three usable DeepSeek keys, with DeepSeek credentials reportedly obtained from servers associated with labibiyat.com and nyisu.com.

    Hackers Use 1,438 Fake FOFA Accounts to Steal OpenAI and DeepSeek API Keys Across the Internet | Cryptika Cybersecurity
  6. Reported Nov 10, 20267d ago

    Attacker automates creation of at least 1,438 FOFA accounts

    An unidentified attacker used a private catch-all email system on melocc.cc to register and activate at least 1,438 FOFA accounts automatically. The attacker rotated their tokens to bypass per-account search limits and obtain tens of thousands of free searches daily.

    Hackers Use 1,438 Fake FOFA Accounts to Steal OpenAI and DeepSeek API Keys Across the Internet | Cryptika Cybersecurity
Linked entities

Related entities

Vulnerabilities, threat actors, malware, products, organizations, breaches, and observables Mallory has linked to this story. Indicator values are masked here and available in full in the app.

Open in app →

Organizations

3 linked
DeepseekOpenaiShanghai Publishing and Printing College

Observables

11 linked

Indicator values are masked on this page. View all 11 in Mallory Domains, IPs, hashes, and URLs are exportable to your SIEM.

Do you run this?See attribution, scope, and whether this vendor sits anywhere in your supply chain.
Get the Full Incident Picture
Mallory platform

See the full picture, correlated to your attack surface.

This page covers what’s public. Mallory adds the parts that aren’t: which of your assets are affected, which threat actors are using it right now, and what to deploy next.

Exposure mappingMap indicators from this story to your assets and identify affected systems in minutes.
Threat actor evidenceEvery observed campaign, victim, and pivot linked to actors named in this story.
Associated malwareMalware, exploits, and IOCs connected to the activity described here.
Detection signaturesYARA, Sigma, and Snort rules deployed to your SIEM as soon as they’re published.
Scheduled alertsGet matching new stories delivered to your team as they break, not the next morning.
AI threadsAsk questions about this story and take action on the answers.