# StepFun

> As of 2026-09-09, StepFun is a threat actor tracked by Threadlinqs Intelligence across 2 threats spanning threat intel, apt. ATT&CK coverage spans 18 techniques across 10 tactics in 2 of 2 tracked threats. Most-observed techniques: AML.T0008 (Acquire Infrastructure), AML.T0024.002 (Exfiltration via AI Inference API: Extract AI Model), AML.T0040 (AI Model Inference API Access).

- **Tracked threats:** 2
- **Categories:** THREAT_INTEL, APT
- **As of:** 2026-09-09

## ATT&CK techniques observed

18 techniques observed across 2 of 2 tracked threats. Tactics: Stealth (formerly Defense Evasion) (4), Exfiltration (3), Resource Development (3), Command and Control (2), Impact (2), Initial Access (2).

- AML.T0008 Acquire Infrastructure — Resource Development — observed in 2 of 2 tracked threats
- AML.T0024.002 Exfiltration via AI Inference API: Extract AI Model — Exfiltration — observed in 2 of 2 tracked threats
- AML.T0040 AI Model Inference API Access — AI Model Access (ATLAS) — observed in 2 of 2 tracked threats
- AML.T0042 Verify Attack — AI Attack Staging (ATLAS) — observed in 2 of 2 tracked threats
- [AML.T0051](https://intel.threadlinqs.com/technique/AML.T0051) LLM Prompt Injection — Execution — observed in 2 of 2 tracked threats
- [AML.T0054](https://intel.threadlinqs.com/technique/AML.T0054) LLM Jailbreak — Stealth (formerly Defense Evasion) — observed in 2 of 2 tracked threats
- [T1048](https://intel.threadlinqs.com/technique/T1048) Exfiltration Over Alternative Protocol — Exfiltration — observed in 2 of 2 tracked threats
- [T1078](https://intel.threadlinqs.com/technique/T1078) Valid Accounts — Initial Access — observed in 2 of 2 tracked threats
- [T1583.003](https://intel.threadlinqs.com/technique/T1583.003) Virtual Private Server — Resource Development — observed in 2 of 2 tracked threats
- AML.T0024 Exfiltration via AI Inference API — Exfiltration — observed in 1 of 2 tracked threats
- AML.T0048 External Harms — Impact — observed in 1 of 2 tracked threats
- AML.T0048.004 External Harms: AI Intellectual Property Theft — Impact — observed in 1 of 2 tracked threats
- [T1027](https://intel.threadlinqs.com/technique/T1027) Obfuscated Files or Information — Stealth (formerly Defense Evasion) — observed in 1 of 2 tracked threats
- [T1036](https://intel.threadlinqs.com/technique/T1036) Masquerading — Stealth (formerly Defense Evasion) — observed in 1 of 2 tracked threats
- [T1090.003](https://intel.threadlinqs.com/technique/T1090.003) Multi-hop Proxy — Command and Control — observed in 1 of 2 tracked threats

## Tracked threats

- [China-Based AI Companies Conducting Industrial-Scale Distillation Campaigns Against U.S. Frontier AI Models](https://intel.threadlinqs.com/threat/TL-2026-2413) — CRITICAL
- [China-Based AI Companies Conducting Industrial-Scale Knowledge Distillation Campaigns Against U.S. Frontier AI Models](https://intel.threadlinqs.com/threat/TL-2026-2405) — HIGH

## Full data

Infrastructure, IOC values and detection queries (Splunk SPL / Microsoft KQL / Sigma) require the Threadlinqs MCP server (Purple tier): https://intel.threadlinqs.com/mcp

Canonical: https://intel.threadlinqs.com/actor/StepFun