Reported by 7 outlets
x47.c Windows Botnet Weaponizes xAI Grok, AI API Draining
Why this score
- Breadth
- 2.00 / 2.0 · 7 sources
- Authority
- 1.80 / 2.0 · best outlet weight
- Severity
- 0.90 / 5.0 · CVSS + content signals
- Velocity
- 0.00 / 2.0 · updates in 48h
- Pickup speed
- 0.33 / 0.6 · time to second source
- Recency
- × 0.00 temporal multiplier
- Community
- + 0.00 / 1.5 · reddit signal
- Score
- 0.0 / 10 = base × recency + community
Desk analysis
Qrator Labs published its write-up of x47.c on September 23, three days before the story surfaced here, and the seller's own advertisement dates to August 3. x47.c is a Windows botnet sold by the actor WraithTools at $200 for the base package, $150 for the DDoS add-on, and $950 for the full set with credential theft, SOCKS5 proxying, and AI-assisted persistence. The panel exposes 18 DDoS methods, from HTTP floods, slow HTTP, TCP and UDP floods, and a TLS stresser to reflection and amplification. Two modules set it apart: an AI API drain that takes a model name and a valid key for the target's OpenAI, xAI, or compatible chat account and bills it directly at the provider, and an AI Stealth module that queries xAI's Grok to pick persistence and concealment actions on the infected host, including Windows Defender exclusions, with local fallbacks when model calls fail. The assessment rests on seller advertisements, technical documentation, and panel screenshots; no confirmed infections accompany it.
The financial module is the threat, not the DDoS menu. OWASP calls this pattern denial of wallet, and its defining property is direction of travel: billable requests leave the victim's environment straight for the provider, so WAF rules, web logs, and application-side controls see nothing while the site stays up and the credits behind it burn. The seller pitched the drain against chatbots, AI-connected CMS platforms, trading bots, and competitors, and suggested automatic top-ups to keep charges accruing after a balance hits zero, which turns a capped nuisance into uncapped loss. The stealer already lists AI-site tokens among its targets even though the documentation stops short of showing them converted into drain keys; that gap is one script wide. There is no CVE and no patch cycle here, so the entire control surface is key hygiene and provider-side spend telemetry, which most organizations do not watch the way they watch their own logs.
Concretely: put hard spend caps and low per-key quotas on OpenAI and xAI accounts, disable automatic credit top-ups, and alert on anomalous token consumption in the provider dashboards, because your own infrastructure will not see the requests. Restrict egress to AI API endpoints to the specific hosts that need it, so a compromised server cannot bill your keys. On endpoints, audit Windows Defender exclusion lists for unapproved entries, watch for browser password, cookie, and Discord token theft, and treat unexplained outbound SOCKS5 relaying as a compromise signal. When triaging the botnet's fast-flux domains, note that several can resolve to a single server, so pivot on IP before sizing the infrastructure.
Two developments would move this assessment. Independent telemetry of real x47.c infections, or a victim report of a drain executed through it, would turn a capability sale into an active campaign. Proof that the stealer's harvested AI tokens are being converted into working API keys would fuse the credential theft and billing chains into one pipeline; precedent for the loss side already exists, since earlier reporting documents a stolen METR API key used to burn roughly $600,000 in AI credits. Qrator found no test results behind the advertised protection-bypass modes, so the claimed capabilities deserve skepticism until demonstrated.
Timeline
-
original reporting
x47.c Windows Botnet Weaponizes xAI Grok, AI API Draining
securityweek.com · 17d ago · September 26th, 2026
A threat actor has been selling access to a new Windows botnet that relies on AI to maintain persistence on infected hosts, Qrator reports. Dubbed x47.c , it is advertised by a threat actor named WraithTools as providing distributed denial-of-service (DDoS) and credential theft capabilities, SOCKS5 proxies, and an AI API drain method. In early August, the threat actor offered the botnet’s base package for $200 and asked $150 for the DDoS add-on. The entire x47.c package, containing the full botnet capabilities, was priced at $950. Users are provided with a command-and-control (C&C) panel that provides access to bot management, fast-flux configuration options, information stealer logs, proxies, concealment capabilities, and DDoS attack options. The panel’s DDoS tab provides access to 18 attack methods, including HTTP floods, AI API draining, slow HTTP, TCP and UDP floods, TLS stresser, and various reflection and amplification techniques. While the HTTP, TCP, and UDP methods are meant to exhaust resources and bandwidth, the AI drain mode is meant to consume a victim’s paid AI credits. The operator supplies a model name and a valid API key for the targeted account on OpenAI, xAI, and compatible chat APIs to consume credits or incur charges. “Because those requests go straight to the provider, they do not need to pass through the victim’s application. The website can remain reachable while the account behind its AI features runs out of c
-
update from qrator.net
x47.c botnet comes with 18 attack methods, including AI API draining
qrator.net · 21d ago · September 23rd, 2026
Original disclosure by Qrator Research: x47.c is a previously undocumented Windows botnet advertised by WraithTools with 18 attack methods, credential theft, SOCKS5 proxies, fast-flux C2, and an AI module meant to keep it on infected machines. Publication date of September 23, 2026 is corroborated by same-day Infosecurity Magazine coverage citing Qrator; the Qrator page itself was not directly fetchable due to bot protection.
-
update from infosecurity-magazine.com
Windows Botnet x47.c Offers AI API Draining, 18 Attack Methods
infosecurity-magazine.com · 20d ago · September 23rd, 2026
Details the AI API drain against OpenAI, xAI, or compatible chat APIs as an OWASP denial-of-wallet attack, and describes an AI Stealth module using xAI's Grok to choose persistence actions including Windows Defender exclusions. Reports an August 3 advertisement pricing the botnet from $200 to $950 and notes Qrator found no test results backing the advertised protection-bypass modes.
-
update from hazetec.com
x47.c Windows Botnet: AI API Credit Draining and 18 Attack Methods Revealed
hazetec.com · 20d ago · September 23rd, 2026
Confirms the story was in public circulation three days before the story's first-seen timestamp.
-
update from cybernews.com
AI API drain botnet can run up huge bills for victims
cybernews.com · 20d ago · September 24th, 2026
Cybernews reports that Qrator Labs found x47.c, a Windows botnet package advertising an AI API drain feature in which valid API keys are used to send billable AI requests and exhaust victims' credits. Article date of September 24, 2026 is taken from Cybernews's own X post sharing the piece (page itself is Cloudflare-challenged).
-
update from scworld.com
New Windows botnet offers AI credit draining and other attack methods
scworld.com · 19d ago · September 24th, 2026
SC World brief noting the newly identified x47.c botnet offers 18 attack methods, including one designed to deplete victims' paid AI credits.
-
mentioned by radar.offseq.com
x47.c Windows botnet (HIGH) weaponizes xAI Grok for persistence, DDoS, credential theft, and AI API draining. Fast-flux C2 & rootkit modules
radar.offseq.com · 17d ago · September 26th, 2026