Back to Intelligence

BAE Systems Cyber Incident Shuts Down US Bankruptcy Noticing Center: Third-Party Risk and Detection Playbook for Defenders

SA
Security Arsenal Team
October 18, 2026
10 min read

On October 9, a security incident at BAE Systems Inc. — the US subsidiary of the British defense giant — knocked the Bankruptcy Noticing Center (BNC) offline. For a full week, every bankruptcy court in the United States lost the ability to send notices to creditors, trustees, attorneys, and debtors. This is not a niche outage. The BNC is the centralized noticing backbone of the federal judiciary, processing millions of notices annually. When it goes dark, court deadlines, creditor rights, and due-process obligations all hang in the balance.

As of this writing, BAE Systems has disclosed few technical details: no confirmed attack vector, no named threat actor, no CVE, and no public confirmation of data exfiltration. What we do know is operationally significant — a single contractor incident cascaded into a nationwide disruption of a judicial function. That is the headline defenders should internalize: your organization's blast radius is defined by your vendors' security posture, not your own.

This post breaks down what happened, what the incident tells us about the state of third-party compromise in 2025–2026, and — most importantly — what your SOC should be hunting for and hardening today.

Technical Analysis: What We Know and What It Implies

The Affected Service

The Bankruptcy Noticing Center, operated under contract for the Administrative Office of the US Courts, handles the generation and distribution of bankruptcy case notices. BAE Systems Inc. holds the contract supporting this function. The service went offline on October 9 and remained unavailable for approximately one week, forcing courts to fall back on manual noticing procedures and delaying time-sensitive legal notifications.

The Attack Surface: Why Contractors Are the Soft Underbelly

No CVE or exploitation vector has been publicly attributed to this incident, and we will not speculate on a specific vulnerability. However, the pattern — a week-long outage at a major defense contractor's subsidiary — is consistent with the intrusion profiles we respond to most often in 2025 and 2026:

  • Ransomware or destructive malware requiring full environment rebuilds (week-long outages are characteristic of detonation-and-restore scenarios, not simple patching)
  • Compromised remote access infrastructure — VPN concentrators, RMM tools, and remote support gateways remain the top initial access vectors in our IR caseload
  • Identity compromise — credential theft via infostealers and adversary-in-the-middle phishing kits, followed by lateral movement into production systems
  • Living-off-the-land post-exploitation — attackers using legitimate administrative tooling (PsExec, WMI, PowerShell remoting) to move from the corporate network into service-delivery infrastructure

The MITRE ATT&CK techniques most relevant to this class of incident include T1078 (Valid Accounts), T1133 (External Remote Services), T1486 (Data Encrypted for Impact), and T1498/T1499 (Denial of Service via service disruption).

Exploitation Status

There is no public proof-of-concept, no CISA KEV entry, and no confirmed threat actor attribution associated with this incident at the time of writing. Treat this as an unattributed intrusion with confirmed operational impact. The absence of attribution does not lower the severity — it raises the uncertainty, which is its own risk.

Detection & Response

Because the specific intrusion vector is undisclosed, the detections below target the generic-but-high-fidelity behaviors we consistently observe in contractor intrusions that result in week-long outages: unauthorized remote access tooling, mass service disruption, and abnormal service account authentication. These are tuned to be useful, not noisy.

SIGMA Rules

YAML
---
title: Unauthorized Remote Monitoring and Management Tool Execution
id: 3b7a9c41-2e5d-4f8a-b912-6c4d8e1f5a07
status: experimental
description: Detects execution of remote access tools frequently abused in contractor intrusions and ransomware precursor activity. Tune the allowlist to your approved RMM stack.
references:
  - https://attack.mitre.org/techniques/T1219/
  - https://securityaffairs.com/201166/security/bae-systems-cyber-incident-takes-us-bankruptcy-noticing-center-offline-for-a-week.html
author: Security Arsenal
date: 2026/01/15
tags:
  - attack.command_and_control
  - attack.t1219
logsource:
  category: process_creation
  product: windows
detection:
  selection:
    Image|endswith:
      - '\anydesk.exe'
      - '\teamviewer.exe'
      - '\teamviewer_service.exe'
      - '\screenconnect.exe'
      - '\splashtop.exe'
      - '\aeroadmin.exe'
      - '\rustdesk.exe'
      - '\meshagent.exe'
      - '\ngrok.exe'
  filter_approved_rmm:
    Image|startswith:
      - 'C:\Program Files\ApprovedVendor\'
  condition: selection and not filter_approved_rmm
falsepositives:
  - Approved IT remote support tooling not yet in the allowlist
level: high
---
title: Mass Service Stop or Disable Activity
tid: 8f2c5d93-7a1b-4e6c-9d03-2b8f4a6e1c59
status: experimental
description: Detects bulk stopping or disabling of Windows services, a hallmark of ransomware pre-encryption staging and destructive incidents that cause prolonged outages.
references:
  - https://attack.mitre.org/techniques/T1489/
author: Security Arsenal
date: 2026/01/15
tags:
  - attack.impact
  - attack.t1489
logsource:
  category: process_creation
  product: windows
detection:
  selection_sc:
    Image|endswith:
      - '\sc.exe'
      - '\net.exe'
      - '\net1.exe'
    CommandLine|contains:
      - ' stop '
      - ' config '
      - ' delete '
  selection_disabled:
    CommandLine|contains:
      - 'start= disabled'
      - 'start=disabled'
  selection_ps:
    CommandLine|contains:
      - 'Stop-Service'
      - 'Set-Service'
      - 'Disable-Service'
  condition: 1 of selection_*
falsepositives:
  - Automated patching and maintenance windows — correlate with change tickets
  - Legitimate administrator service management on a single host
level: medium
---
title: Backup Catalog and Shadow Copy Deletion Attempt
tid: 1d4e7b26-9c3a-4f58-a672-5e9b2c8d3f41
status: experimental
description: Detects deletion of volume shadow copies and backup catalogs, a common precursor to outages requiring full environment rebuilds such as week-long service disruptions.
references:
  - https://attack.mitre.org/techniques/T1490/
author: Security Arsenal
date: 2026/01/15
tags:
  - attack.impact
  - attack.t1490
logsource:
  category: process_creation
  product: windows
detection:
  selection_vss:
    Image|endswith: '\vssadmin.exe'
    CommandLine|contains:
      - 'delete shadows'
      - 'Delete Shadows'
  selection_bcdedit:
    Image|endswith: '\bcdedit.exe'
    CommandLine|contains:
      - 'recoveryenabled'
      - 'bootstatuspolicy'
  selection_wbcdmin:
    Image|endswith: '\wbadmin.exe'
    CommandLine|contains:
      - 'delete catalog'
      - 'delete systemstatebackup'
  condition: 1 of selection_*
falsepositives:
  - Backup software performing catalog maintenance (verify process ancestry)
level: critical

KQL — Microsoft Sentinel / Defender

The following hunt query surfaces anomalous authentication and process behavior for service accounts and third-party/vendor accounts — the identities most commonly abused in contractor intrusions. Run it against a 14-day lookback and compare against each account's historical baseline.

KQL — Microsoft Sentinel / Defender
// Hunt: Anomalous logons and tool execution by service/vendor accounts
let Lookback = 14d;
let ServiceAccounts = dynamic(["svc-", "vendor", "contractor", "rmm", "support-"]);
union isfuzzy=true
(DeviceProcessEvents
| where TimeGenerated > ago(Lookback)
| where AccountName has_any (ServiceAccounts)
| where FileName in~ ("psexec.exe","wmic.exe","sc.exe","net.exe","net1.exe","vssadmin.exe","bcdedit.exe","wbadmin.exe","powershell.exe","pwsh.exe","rundll32.exe","regsvr32.exe","msiexec.exe","anydesk.exe","teamviewer.exe","screenconnect.exe")
| project TimeGenerated, DeviceName, AccountName, FileName, ProcessCommandLine, InitiatingProcessFileName, InitiatingProcessCommandLine
),
(SigninLogs
| where TimeGenerated > ago(Lookback)
| where UserPrincipalName has_any (ServiceAccounts)
| summarize SignInCount=count(), DistinctIPs=dcount(IPAddress), IPs=make_set(IPAddress, 10), Countries=make_set(Location, 10) by UserPrincipalName, bin(TimeGenerated, 1h)
| where DistinctIPs > 3 or SignInCount > 50
);
// Secondary: single host authenticating to many targets in short window (lateral movement)
let Window = 10m;
let Threshold = 15;
DeviceNetworkEvents
| where TimeGenerated > ago(24h)
| where RemotePort in (445, 3389, 5985, 5986, 22)
| summarize Targets=dcount(RemoteIP), TargetList=make_set(RemoteIP, 20) by DeviceName, InitiatingProcessFileName, bin(TimeGenerated, Window)
| where Targets >= Threshold
| sort by Targets desc;

Velociraptor VQL

Use this artifact to sweep endpoints for unapproved RMM tooling and shadow copy tampering evidence — the two artifacts most often left behind in incidents that produce multi-day outages.

VQL — Velociraptor
-- Hunt for unauthorized RMM binaries and shadow copy tampering artifacts
LET rmm_paths = SELECT FullPath, Size, Mtime
FROM glob(globs=[
  'C:/Users/*/AppData/**/AnyDesk.exe',
  'C:/Users/*/AppData/**/rustdesk.exe',
  'C:/Users/*/Downloads/**/ScreenConnect*.exe',
  'C:/ProgramData/**/AnyDesk.exe',
  'C:/Program Files*/TeamViewer/TeamViewer*.exe',
  'C:/Program Files*/Splashtop/**/SRManager.exe'
])
WHERE NOT FullPath =~ 'ApprovedVendor';

LET suspicious_procs = SELECT Pid, Name, CommandLine, Exe, Username, CreateTime
FROM pslist()
WHERE CommandLine =~ '(?i)(delete shadows|recoveryenabled no|delete catalog|stop-service|net stop)'
   OR Name =~ '(?i)(anydesk|rustdesk|screenconnect|teamviewer)';

SELECT * FROM rmm_paths
UNION ALL
SELECT FullPath=NULL, Size=NULL, Mtime=NULL, Pid, Name, CommandLine, Exe, Username, CreateTime FROM suspicious_procs;

Verification & Hardening Script

This PowerShell script audits a Windows host for the most common intrusion artifacts in contractor-breach scenarios: unapproved RMM software, disabled services, tampered shadow copies, and recently created local accounts. Run it via your EDR or configuration management tooling across the fleet.

PowerShell
# Security Arsenal — Third-Party Intrusion Artifact Audit
# Run elevated. Review output before taking remediation action.

$Report = @()

# 1. Enumerate installed RMM tools (compare against your approved list)
$RMMNames = 'AnyDesk','TeamViewer','ScreenConnect','Splashtop','AeroAdmin','RustDesk','ConnectWise','LogMeIn','GoToMyPC'
$Installed = Get-ItemProperty HKLM:
SOFTWARE\Microsoft\Windows\CurrentVersion\Uninstall\*,
  HKLM:\SOFTWARE\WOW6432Node\Microsoft\Windows\CurrentVersion\Uninstall\* -ErrorAction SilentlyContinue |
  Where-Object { $n = $_.DisplayName; $RMMNames | Where-Object { $n -match $_ } } |
  Select-Object DisplayName, DisplayVersion, InstallDate
$Report += [pscustomobject]@{ Check='RMM Tools Found'; Result = ($Installed | Out-String) }

# 2. Verify Volume Shadow Copies are intact and not recently deleted
$Shadows = Get-WmiObject Win32_ShadowCopy -ErrorAction SilentlyContinue
$Report += [pscustomobject]@{ Check='Shadow Copies Present'; Result = ($null -ne $Shadows) }
$Report += [pscustomobject]@{ Check='VSS Service State'; Result = (Get-Service VSS).Status }

# 3. Detect services recently disabled (potential staging for disruption)
$Disabled = Get-CimInstance Win32_Service | Where-Object { $_.StartMode -eq 'Disabled' -and $_.State -eq 'Stopped' } |
  Select-Object Name, DisplayName, PathName
$Report += [pscustomobject]@{ Check='Disabled Services Count'; Result = $Disabled.Count }
$Disabled | Export-Csv -Path "$env:TEMP\disabled_services.csv" -NoTypeInformation

# 4. Local accounts created in the last 30 days (persistence check)
$NewAccounts = Get-LocalUser | Where-Object { $_.Enabled -and $_.PasswordLastSet -gt (Get-Date).AddDays(-30) }
$Report += [pscustomobject]@{ Check='Recently Enabled Local Accounts'; Result = ($NewAccounts.Name -join ', ') }

# 5. Check recovery configuration (bcdedit tamper indicator)
$Bcd = bcdedit | Out-String
$Report += [pscustomobject]@{ Check='Recovery Enabled'; Result = ($Bcd -match 'recoveryenabled\s+Yes') }

$Report | Format-List
# Remediation guidance:
# - Uninstall any RMM tool not on the approved list and block its binary via AppLocker/WDAC
# - Re-enable shadow copies: vssadmin list shadows / vssadmin create shadow
# - Investigate any disabled critical service against change management records
# - Disable and investigate unauthorized local accounts; force credential rotation

Remediation: What Defenders Should Do Now

Because no CVE or specific patch is associated with this incident, remediation is architectural and procedural rather than a version upgrade. Prioritize the following:

  1. Map your critical-service dependencies. Produce an inventory of every third party whose outage would halt a business-critical function — as BNC did for the judiciary. For each, document the fallback procedure and test it. The courts' fallback to manual noticing was slow and partial; yours should not be.

  2. Enforce contractual security telemetry from vendors. Require evidence of EDR coverage, MFA on all remote access, and 24/7 monitoring from any vendor operating a service on your behalf. If your contracts don't include breach notification SLAs measured in hours (not days), renegotiate.

  3. Harden remote access paths. Disable or tightly control RMM tooling via application allowlisting (AppLocker/WDAC). Require phishing-resistant MFA (FIDO2/passkeys) on VPN, VDI, and any externally reachable administrative interface. Infostealer-harvested credentials remain the dominant initial access vector in 2025–2026.

  4. Deploy the detections above. The Sigma rules, KQL, and VQL in this post target the behaviors that precede week-long outages: shadow copy deletion, mass service disruption, and unauthorized remote access. Tune allowlists to your environment before broad deployment.

  5. Test your ransomware rebuild time. A week of downtime implies a rebuild-from-backup scenario. If your last tabletop exercise didn't include restoring a production service from immutable backups under adversary pressure, schedule one this quarter.

  6. Monitor for follow-on disclosure. Watch the Administrative Office of the US Courts notices, BAE Systems regulatory filings, and CISA advisories for attribution and indicator release. When IOCs drop, hunt retroactively across at least 90 days of telemetry — dwell time in contractor environments routinely exceeds that window.

Related Resources

Security Arsenal Managed SOC Services AlertMonitor Platform Book a SOC Assessment soc-mdr Intel Hub

Is your security operations ready?

Get a free SOC assessment or see how AlertMonitor cuts through alert noise with automated triage.