GreyNoise recorded two distinct campaigns probing exposed large language model (LLM) infrastructure between October 2025 and January 2026. One used server-side request forgery (SSRF) techniques to trigger outbound connections; the other mapped model endpoints and misconfigured proxies. The activity shows how exposed integrations can create security risks, but the reporting does not establish that either campaign stole corporate secrets or breached an organization.
What GreyNoise observed
GreyNoise says its Ollama honeypot infrastructure captured 91,403 attack sessions from October 2025 through January 2026. That is a count of sessions seen by its honeypots—not a count of unique attackers, successful intrusions, or confirmed breaches. The activity divided into two campaigns with different techniques and apparent purposes, according to GreyNoise’s January 8, 2026 report, “Threat Actors Actively Targeting LLMs”.
Dark Reading summarized the findings on January 12, 2026, in “2 Separate Campaigns Probe Corporate LLMs for Secrets”. The phrase “for secrets” describes the risk of exposing LLM infrastructure; the reporting does not document secrets being stolen.
How the two campaigns differed
| Campaign | Observed technique and target surface | Reported scale and timing | GreyNoise’s assessment |
|---|---|---|---|
| SSRF activity | Malicious registry URLs in Ollama model pulls and manipulated MediaUrl parameters in Twilio SMS webhook integrations to induce outbound connections. |
October 2025 through January 2026; 1,688 sessions in a 48-hour Christmas spike. GreyNoise reported 62 source IPs across 27 countries. | Probably security researchers or bug bounty hunters, though the scale and timing suggested gray-hat activity. The operators’ identities are not established. |
| Endpoint enumeration | Repeated probes of LLM model endpoints and misconfigured proxies that could expose access to commercial APIs. | Started December 28, 2025; two IP addresses generated 80,469 sessions in eleven days and probed more than 73 endpoints. | Professional threat-actor reconnaissance, in GreyNoise’s assessment; this is not a confirmed attribution. |
GreyNoise also associated the two enumeration IPs with more than four million historical sensor observations tied to previous exploitation activity. That figure refers to GreyNoise’s combined sensor observations, not to attacks in this campaign.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallOutdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware match#1 Best Overall
- Compact and Efficient Design: The FortiGate 40F is designed for small to mid-sized businesses and enterprise branch offices, featuring a compact, fanless desktop form factor that ensures quiet operation and minimizes space usage.
- Robust Connectivity Options: Equipped with 5 GE RJ45 ports, including 1 WAN port and 4 internal ports, this model provides essential connectivity and flexibility for various network configurations in a small-scale environment.
- High-Performance Security: Offers up to 1 Gbps IPS throughput and 600 Mbps threat protection throughput, using Fortinet’s purpose-built security processor technology to deliver industry-leading performance and protection for SSL encrypted traffic.
- Advanced Threat Protection: Integrated with Fortinet’s AI-powered FortiGuard Labs, the FortiGate 40F offers comprehensive cybersecurity, identifying and mitigating both known and unknown threats to maintain robust security across your network.
- Simplified Management and Deployment: Features a user-friendly management console that provides comprehensive network automation and visibility, coupled with Zero Touch Integration with Fortinet’s Security Fabric for easy deployment.
SSRF: inducing outbound callbacks
The first campaign used Ollama model-pull URLs and Twilio webhook parameters to cause exposed systems to make outbound connections. GreyNoise reported that the activity used ProjectDiscovery’s OAST infrastructure to validate callbacks. It saw a spike of 1,688 sessions in 48 hours around Christmas. GreyNoise assessed the activity as likely security research or bug bounty testing, while noting that its scale and timing suggested gray-hat behavior. The evidence does not identify the operators.
Enumeration: identifying models behind endpoints
The second campaign began December 28, 2025. Two IP addresses sent 80,469 sessions in eleven days, methodically probing more than 73 model endpoints for misconfigured proxies that could expose access to commercial APIs. The prompts were deliberately innocuous; GreyNoise assessed that they were intended to identify which model answered without attracting alerts.
Rank #2
- HARDWARE PLUS SECURITY SERVICES: FortiGate-60F Firewall Appliance bundled with 1 year of FortiCare Premium and FortiGuard Unified Threat Protection.
- UNIFIED THREAT PROTECTION (UTP): Secures against advanced online threats with comprehensive web filtering and anti-botnet technologies.
- OPTIMIZED FOR MEDIUM-SIZED BUSINESSES: Tailored for businesses needing robust security without the infrastructure of larger enterprises.
- RELIABLE CUSTOMER SUPPORT: FortiCare Premium ensures high-quality support and service continuity.
- EFFECTIVE PROTECTION: Employs advanced filtering technologies to safeguard against sophisticated threats.
Reported example prompts included “How many states are there in the United States?” and “What model are you?” GreyNoise also reported the combined prompt “How many states are there in the United States? What is todays date? What model are you?” The probe list covered OpenAI-compatible API formats and Google Gemini formats, and named model families including OpenAI GPT-4o and variants, Anthropic Claude Sonnet, Opus, and Haiku, Meta Llama 3.x, DeepSeek-R1, Google Gemini, Mistral, Alibaba Qwen, and xAI Grok.
What the activity does—and does not—show
The strongest supported conclusion is that exposed LLM endpoints and adjacent integrations faced large-scale probing and attempts to induce outbound connections. Endpoint enumeration can help an operator identify potential targets for later activity, which is why GreyNoise warned that the probing could be preparatory.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Rank #3
- 【Up to 1100 Mbps VPN Speed 】 Hardware-accelerated WireGuard and OpenVPN-DCO deliver up to 1100 Mbps VPN throughput, over 3× faster than Brume 2 for smooth remote access and file transfers.
- 【Three 2.5G Ports & Multi-WAN】Tri-port 2.5GbE design with flexible WAN LAN configuration supports multi-gigabit wired setups, dual-ISP Multi-WAN and failover to keep home and SOHO networks online.
- 【Stealth VPN Obfuscation】VPN obfuscation disguises VPN traffic as regular HTTPS, helping you evade blocking, bypass restrictive networks and maintain stable, private connections.
- 【DPI protection】Deep Packet Inspection with visual dashboards blocks adult/gambling/malicious sites, while SQM and QoS prioritize gaming, calls, and video when bandwidth is tight
- 【OpenWrt & USB 3.0 Expansion】OpenWrt with 1GB DDR4 and 8GB eMMC lets you install plugins and build VPN, ad-blocking or NAS, while USB 3.0 Type‑C connects high-speed storage or 4G/5G dongles
GreyNoise researcher Bob Rudis wrote, “Eighty thousand enumeration requests represent investment.” He also wrote, “Threat actors don’t map infrastructure at this scale without plans to use that map.” Those statements express an interpretation of scale and likely intent; they are not proof of a specific follow-on attack. Neither GreyNoise nor Dark Reading documented successful theft of company secrets from these campaigns.
The campaigns should not be collapsed into one actor or objective: GreyNoise characterized the SSRF activity as likely research or gray-hat testing and the enumeration activity as more concerning reconnaissance. Both are researcher assessments, not confirmed identities or motives.
Rank #4
- Runs UniFi Network for full-stack network management
- Manages 30+ UniFi Network devices and 300+ clients
- 1 Gbps routing with IDS/IPS
- Multi-WAN load balancing
- 0.96" LCM status display
Defensive steps for enterprise teams
GreyNoise recommends controls aimed at the observed techniques. These are risk-reduction measures, not a guarantee of protection. Validate indicators against current telemetry before blocking: IP addresses, autonomous system numbers (ASNs), JA4 fingerprints, and domains can change or be shared by unrelated activity.
- Constrain model sources and outbound traffic. Restrict Ollama model pulls to trusted registries, and apply egress filtering to limit callbacks to attacker-controlled infrastructure.
- Detect endpoint sweeps and fingerprinting. Alert on rapid-fire requests spanning multiple model endpoints, including low-risk prompts that may be used to identify the responding model.
- Control callback resolution. Consider blocking known out-of-band application security testing (OAST) domains at DNS, after validating that doing so will not disrupt legitimate testing.
- Review rate limits and network indicators. Rate-limit suspicious ASNs that appeared prominently in this traffic, and monitor the JA4 fingerprints GreyNoise identified for this tooling and similar automation.
- Check the proxy boundary. Review exposed LLM proxies and integrations for unintended access to commercial API credentials or outbound network paths, then restrict access and egress to what the service requires.
GreyNoise’s recommendations and campaign details are available in its January 8, 2026 report.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
