Skip to navigation

Threat reportThreat IntelligenceTL-2026-1250

Hardware Trojan Backdoors in Chip Design Detected via AI-Assisted Verification (VeriChat)

MONITORING

University of Florida researchers demonstrated a hardware Trojan hidden in an AES S-Box IP block that activates on a specific 3-byte trigger sequence (0xDE, 0xAD, 0xBE) and leaks the AES secret key one bit at a time via a status-light side channel over eight clock cycles, with the trigger firing spuriously only about 6 times per 100 million cycles.

Key facts

IdentifiersTL-2026-1250
Also known asVeriChat AES S-Box Hardware Trojan Demonstration
SeverityINFORMATIONAL
Affected productsAcademic Research Demonstration Synthesizable AES S-Box RTL IP block (research testbed)
Weaknesses (CWE)CWE-507, CWE-1234, CWE-1244, CWE-1245, CWE-1300
MalwareAES S-Box Hardware Trojan (research demonstration)
ToolsIcarus Verilog, SymbiYosys, VeriChat, Yosys
MITRE ATT&CK techniquesT1005 Data from Local System, T1011 Exfiltration Over Other Network Medium, T1027 Obfuscated Files or Information, T1052 Exfiltration Over Physical Medium, T1119 Automated Collection, T1195 Supply Chain Compromise, T1195.003 Compromise Hardware Supply Chain, T1199 Trusted Relationship, T1200 Hardware Additions, T1205 Traffic Signaling, T1495 Firmware Corruption, T1528 Steal Application Access Token and 2 more in the attack flow
MotivationUNKNOWN
Target sectorssemi-conductors, electronics manufacturing, defense industrial base, critical infrastructure, academia
Target regionsNorth America
Detection rules9 (1 shown free)
Sources8 cited sources

Detection coverage

As of 2026-07-13, Threadlinqs Intelligence publishes 9 detection rule(s) for TL-2026-1250 across Splunk SPL, Microsoft KQL and Sigma, covering 15 indicator(s) of compromise. The whole corpus is readable without an account. Every threat comes with one complete detection in Splunk SPL, Microsoft KQL and Sigma, free for everyone; a free account adds daily TL Creds to unlock the rest. Paid plans add a larger daily allowance, correlation and the MCP server. Threadlinqs MCP server · View plans.

Free detection: Trigger-Fired Bit-Serial Status-Light Key Exfiltration Trace — C2/Exfiltration Phase

Detects simulation-based trigger-testing telemetry (VeriChat stage 3 / SymbiYosys stage 4 equivalent) where trigger_fired=true is accompanied by an 8-cycle (or fixed-N-cycle) bit-serial toggle trace on a status/diagnostic output — the exact behavior of the documented AES S-Box Trojan leaking the AES round-key one bit per clock cycle over an ostensibly benign LED. Also covers a formal-verification finding proving a reachable data-flow path from key-storage registers to an externally observable pin.

  • Languages: Splunk SPL, Microsoft KQL, Sigma
  • ATT&CK: T1119, T1052, T1011, T1205, T1528, T1495
  • Data sources: eda_security_scanner_findings_log, symbiyosys_formal_verification_report, verification_pipeline_trigger_simulation_log
  • Severity: critical
  • Confidence: high

Splunk SPL

index=eda_security sourcetype="eda:scanner:findings" trigger_fired=true exfil_trace_simulated="*cycle*"
| eval exfil_cycle_count=mvcount(split(exfil_trace_simulated, "cycle"))-1
| where exfil_cycle_count >= 8
| stats count values(exfil_trace_simulated) as trace by host, rtl_file, module_name
| eval severity="critical"

Microsoft KQL

EDAScannerFindings
| where TimeGenerated > ago(24h)
| where TriggerFired == true
| where ExfilTraceSimulated has "cycle"
| extend CycleCount = array_length(parse_json(ExfilTraceSimulated))
| where CycleCount >= 8
| project TimeGenerated, Computer, RtlFile, ModuleName, CycleCount, ExfilTraceSimulated
| summarize Count=count(), MaxCycles=max(CycleCount) by Computer, RtlFile, ModuleName

Sigma

title: Simulated Trigger-Sequence Firing With Bit-Serial Exfil Trace
id: 5f1e9d2a-1250-4b1a-9c3d-hwtrojan0003
status: experimental
description: Detects a verification-pipeline trigger_fired=true event accompanied by an 8-cycle exfil_trace, consistent with the documented AES S-Box Trojan payload behavior
author: Threadlinqs Team
date: 2026-07-13
tags:
    - attack.collection
    - attack.exfiltration
    - attack.command-and-control
    - attack.t1119
    - attack.t1052
    - attack.t1011
    - attack.t1205
    - attack.t1528
    - attack.t1495
logsource:
    category: application
    product: eda_security_scanner
detection:
    selection:
        trigger_fired: true
    filter:
        exfil_trace_simulated|contains: 'cycle'
    condition: selection and filter
falsepositives:
    - Legitimate self-test/BIST sequences that intentionally emit a fixed-length toggle pattern on a status pin during power-on diagnostics
level: critical

Known false positives:

  • Legitimate power-on self-test (POST/BIST) sequences that emit a fixed-length toggle pattern on a status pin for approved diagnostic purposes

Tuning notes: Treat as critical only when paired with a stage-2 unexplained-state finding (TL-2026-1250-*-002) and a stage-1 magic-byte trigger finding (TL-2026-1250-*-001); a lone trigger_fired event without corroborating synthesis findings should be downgraded to high and manually triaged.

8 more detection rules (Splunk SPL, Microsoft KQL, Sigma) · unlock with TL Creds. Sign up free to unlock the detection rules or query them with the Threadlinqs MCP server (Purple)
  • RTL Magic-Byte Trojan Trigger Constant Detected (0xDEADBE) — Exploitation Phase (KQL) · unlock with TL Creds
  • Unexplained Sequential State Driving Diagnostic/Status Output — Post-Exploitation Phase (KQL) · unlock with TL Creds
  • Trigger-Fired Bit-Serial Status-Light Key Exfiltration Trace — C2/Exfiltration Phase (KQL) · unlock with TL Creds
  • RTL Magic-Byte Trojan Trigger Constant Detected (0xDEADBE) — Exploitation Phase (Sigma) · unlock with TL Creds
  • Unexplained Sequential State Driving Diagnostic/Status Output — Post-Exploitation Phase (Sigma) · unlock with TL Creds
  • Trigger-Fired Bit-Serial Status-Light Key Exfiltration Trace — C2/Exfiltration Phase (Sigma) · unlock with TL Creds
  • RTL Magic-Byte Trojan Trigger Constant Detected (0xDEADBE) — Exploitation Phase · unlock with TL Creds
  • Unexplained Sequential State Driving Diagnostic/Status Output — Post-Exploitation Phase · unlock with TL Creds

Indicators of compromise

As of 2026-07-13, Threadlinqs Intelligence tracks 15 indicators of compromise for TL-2026-1250.

15 indicators of compromise · unlock with TL Creds. Compare plans

Technical analysis

The team built VeriChat, a retrieval-augmented, three-agent conversational AI assistant trained on a curated library of 28,221 hardware security papers and integrated with open-source EDA tools (Icarus Verilog, Yosys, SymbiYosys), to autonomously identify, simulate, and formally prove such covert key-leakage vulnerabilities, achieving 87.73% factual (faithfulness) accuracy and a 92% false-premise rejection rate.

This is an academic hardware-security research demonstration, not an in-the-wild exploited vulnerability. Researchers at the University of Florida (Dipayan Saha, Khan Thamid Hasan, Shams Tarek, Sujan Kumar Saha, Mark Tehranipoor, Farimah Farahmandi) published 'VeriChat: An Agentic Conversational AI Assistant for Hardware Security Verification' (arXiv:2607.01668, submitted 2026-07-02, accepted for presentation at IEEE COINS 2026). The paper's centerpiece case study is a hardware Trojan implanted in a synthesizable AES S-Box RTL IP block: a sequential trigger circuit continuously monitors an input/control bus for the exact 3-byte pattern 0xDE, 0xAD, 0xBE. Once that trigger sequence is observed, a payload state machine begins exfiltrating the AES round-key material one bit at a time by toggling an otherwise-benign status/diagnostic LED output over eight consecutive clock cycles, allowing a physically-proximate or optically-instrumented observer to reconstruct the secret key without any digital output path or overt communication channel. The trigger's false-activation rate on random/benign traffic is characterized as roughly 6 in 100,000,000 cycles, making it statistically invisible to conventional functional and random-pattern verification. To detect this class of Trojan, the researchers built VeriChat: a retrieval-first, three-agent LLM pipeline (question reformulation agent, evidence-gathering agent, answer-generation agent) grounded in a curated corpus of 28,221 hardware-security papers plus live web retrieval, explicitly designed to minimize hallucination and maintain traceable evidence citations. VeriChat is wired into a four-stage automated verification pipeline over the target RTL: (1) syntax verification (Verilog/RTL compiles cleanly via Icarus Verilog), (2) synthesis analysis (Yosys-based structural/memory-element counting to flag unexplained state), (3) simulation-based trigger-sequence testing (driving the exact suspected trigger inputs and observing payload behavior), and (4) formal verification (SymbiYosys-based mathematical proof that the key-bit leakage path exists and is reachable). Expert human review scored VeriChat's factual accuracy at 87.73%, and a false-claim/false-premise rejection test (probing the tool with invented, nonexistent hardware-security concepts such as 'Metamaterial Resonance Shielding') showed a 92% correct-refusal rate, both reported as outperforming leading proprietary general-purpose LLMs on the same evaluation. This threat is retained by the harness as a supply-chain/hardware-trust research signal: it has no CVE, no shipping commercial product, and no observed in-the-wild exploitation, but it is directly relevant to defenders and hardware security teams evaluating third-party silicon IP blocks, chip supply-chain integrity, and next-generation AI-assisted RTL/Trojan-detection tooling.

MITRE ATT&CK techniques by tactic:

Collection

T1005 Data from Local System; T1119 Automated Collection

Exfiltration

T1011 Exfiltration Over Other Network Medium; T1052 Exfiltration Over Physical Medium

Defense Evasion

T1027 Obfuscated Files or Information

initial-access

T1195 Supply Chain Compromise; T1195.003 Compromise Hardware Supply Chain

Initial Access

T1195 Supply Chain Compromise; T1199 Trusted Relationship; T1200 Hardware Additions

Command and Control

T1205 Traffic Signaling

Impact

T1495 Firmware Corruption

Credential Access

T1528 Steal Application Access Token

defense-impairment

T1601 Modify System Image; T1601.001 Patch System Image

Remediation

Immediate actions

  • Treat third-party/vendor RTL and hard IP blocks (especially crypto cores like AES S-Box implementations) as untrusted supply-chain artifacts requiring independent verification before tapeout or FPGA deployment
  • Run suspicious sequential triggers against known 'magic byte' patterns (e.g. 0xDEADBEEF-style constants) during pre-silicon security review
  • Audit any 'diagnostic', 'status', or 'debug' output pins (LEDs, GPIOs, JTAG) on cryptographic IP for unexpected data-dependent toggling behavior

Workarounds

  • Where third-party IP cannot be independently verified, physically isolate or shield diagnostic/status output pins on cryptographic modules from external observation

Longer-term hardening

  • Integrate AI-assisted retrieval-augmented verification tools (e.g. VeriChat-style syntax/synthesis/simulation/formal-verification pipelines) into standard pre-tapeout hardware security sign-off
  • Adopt formal verification (e.g. SymbiYosys) as a mandatory gate for licensed third-party crypto IP, not just functional simulation
  • Establish an internal curated corpus of hardware-Trojan case studies (e.g. Trust-Hub benchmark suite) to train/validate in-house detection tooling
  • Require IP vendors to provide golden netlists and formal non-interference proofs for security-critical blocks

Affected and fixed versions

  • Academic Research Demonstration — Synthesizable AES S-Box RTL IP block (research testbed)
    Vulnerable versions: Trojan-implanted research IP used as VeriChat case study
    Fixed in: N/A — proof-of-concept research artifact, not a shipping product

Timeline

  • University of Florida research team submits 'VeriChat: An Agentic Conversational AI Assistant for Hardware Security Verification' to arXiv (2607.01668).
  • MITRE ATT&CK mapping expanded across Resource Development, Initial Access, Persistence, Defense Evasion, Credential Access, Collection, Exfiltration, Command and Control, and Impact tactics to comprehensively model the hardware-Trojan supply-chain threat, including the T1199 trusted-relationship IP-licensing vector, T1027 trigger-logic obfuscation, and T1205 traffic-signaling-style covert activation.
  • Threat mapped to CWE-507 (Trojan Horse) and CWE-1234/1244/1245/1300 (hardware debug/lock-override and side-channel exposure weaknesses) based on the trigger-and-exfiltration mechanism described in arXiv:2607.01668.
  • VeriChat's four-stage pipeline (syntax verification via Icarus Verilog, synthesis analysis via Yosys, simulation-based trigger testing, and formal verification via SymbiYosys) is documented, alongside its 87.73% factual accuracy and 92% false-premise rejection rate.
  • Analysis confirms the AES S-Box Trojan uses a 3-byte (0xDE, 0xAD, 0xBE) sequential trigger and an 8-cycle status-light bit-serial key exfiltration payload, with a ~6-in-100,000,000-cycle false trigger rate.
  • TL-Intel Harness RESEARCH phase begins deep-dive analysis of the arXiv paper, CWE mappings, and MITRE ATT&CK context.
  • TL-Intel Harness HUNT phase ingests the Help Net Security article via RSS feed monitoring and creates threat skeleton TL-2026-1250.
  • Help Net Security publishes coverage of the VeriChat research and its AES S-Box hardware Trojan case study.

Sources cited

Previous and next in threat intel

Drafted by the Threadlinqs research pipeline from 8 sources; published ; last updated —.

Threadlinqs Intelligence — Real-Time Threat Detection Platform

// threat_feed
$ sort --newest
Showing all threats

Live intelligence console

Threat level

Latest Threats