business

Verdict

Submitted 5/20/2026, 11:56:56 AM · Completed 5/20/2026, 11:58:37 AM

6.2
pivot
The idea

Windows File Server Audit – How to reliably detect who created a file? (EventID 4663 / 5145 / 4656 confusion)

Pain point
The user cannot reliably determine who created a file on a Windows File Server using native audit events.
Who has this problem
Sysadmins managing Windows File Server environments
Contradiction (TRIZ)
Need to distinguish file creation from access while relying solely on event logs which lack deterministic indicators.
Ideal final result
A system that automatically and accurately identifies file creation events with user attribution without requiring additional monitoring tools.
Suggested solution
Implement Sysmon (Event ID 11) for precise file creation tracking, combined with correlation of 4663/5145 events to validate activity patterns.
Show original source text →
Hi everyone, I’m trying to reliably track **file creation activities** on a Windows File Server environment (e.g. “user X created file Y”). # 🔧 Environment * Windows File Server (SMB share) * SIEM integrated (logs forwarded) * Object auditing enabled # 🎯 Goal I want to answer questions like: * *Which user created a file?* * *Which user copied/uploaded a file to the share?* * *Distinguish between file creation vs file access* # 🔍 What I did so far I enabled file auditing (SACL) on the target directory and started collecting these events: * **4663** – An attempt was made to access an object * **4656** – A handle to an object was requested * **5145** – A network share object was checked (Detailed File Share) I can see all these events in my SIEM. # ❗ Problem I’m **not able to reliably identify file creation events**. The issue is: * When a file is **created** * When a file is just **accessed/read** * When metadata is queried (Explorer behavior) 👉 I often see **the same EventIDs and very similar AccessMask values** For example: * `0x100080` (ReadAttributes + SYNCHRONIZE) appears during create **and** access * `0x120089` (read-heavy mask) appears in normal file opening * Even multiple 4663 events look identical between open vs create scenarios So: > # 🔁 Observed Behavior When creating a file: * First I see events with **ReadAttributes (0x100080)** * Then possibly multiple access events * Sometimes I don’t clearly see a strong “write” signature tied directly to the final file When opening an existing file: * I get very similar sequences 👉 Which makes detection ambiguous # ❓ Question Is there a **reliable way (using native Windows Security logs)** to: 1. Detect **file creation events only** 2. Identify the **user who created or uploaded the file** 3. Distinguish: * create vs open * create vs overwrite * copy vs normal access # 🤔 Things I’m considering * Filtering **4663 Accesses for WriteData / AppendData** * Correlating 4663 + 5145 within a time window * Tracking **first-seen filenames** But none of these feel **100% deterministic** # 💬 What I’m looking for * Best practice from real-world environments * Detection logic used in SIEM (KQL / SPL / etc.) * Whether this is even **reliably possible with only 4663/5145/4656** * Or if I should move to something like: * Sysmon (Event ID 11 – FileCreate) * FSRM / File Integrity Monitoring tools # 🚀 Summary 👉 I can see file activity 👉 But I **cannot confidently say “this user created this file” using native audit events only”** Any insights or field-proven approaches would be greatly appreciated 🙏
TRIZ inventive level: 3/5· Principles: parameter changes, correlation
Synthesis verdict
**Pivot**. The idea of reliably tracking file creation activities on a Windows File Server environment using native Windows Security logs is challenging due to the similarity in event IDs and AccessMask values between file creation and access. While there is a real, paying market for this solution, primarily among enterprise IT security teams, compliance officers, and forensic investigators, the technical challenge is significant. The proposed approaches, such as filtering 4663 accesses for WriteData/AppendData or correlating multiple event IDs, show promise but are not 100% deterministic. Alternative solutions like using Sysmon or FSRM/File Integrity Monitoring tools could provide more reliable results. To pivot, the solution should focus on combining native logs with Sysmon/FSRM data to deliver high-confidence creation attribution.

Strengths

  • Real, paying market for the solution, primarily among enterprise IT security teams, compliance officers, and forensic investigators
  • Unit economics are favorable, with low marginal cost per monitored server and high gross margins
  • Opportunity to package the solution as a pre-built SIEM detection pack (KQL/SPL) with tuning guides

Weaknesses

  • Native Windows audit logs cannot reliably distinguish file creation from access
  • The proposed approaches are not 100% deterministic and may be fragile
  • Dependence on Windows native auditing is a single point of failure vulnerable to regulatory, platform, and budget pressures

Best angle

The solution should focus on combining native logs with Sysmon/FSRM data to deliver high-confidence creation attribution, targeting enterprise IT security teams, compliance officers, and forensic investigators.

Panel verdicts

Competition

nvidia/nemotron-3-nano-omni-30b-a3b-reasoning

3.0

Native Windows audit logs cannot reliably distinguish file creation from other access events; a dedicated creation event (e.g., Sysmon ID 11) or FIM is required for a durable solution.

The proposed solution relies solely on Windows Security Event IDs 4663, 4656, and 5145, which all report generic object access attempts. These events share overlapping AccessMask values (e.g., 0x100080) and lack a definitive write‑specific flag that unambiguously marks a creation event versus a read or metadata query. Real‑world deployments show that the same event patterns appear for opening an existing file, copying data, or merely querying attributes, making it impossible to reliably infer "file created" without additional context. While correlating 4663 with 5145 or filtering for AppendData may improve confidence, the approach remains probabilistic and fragile; any change in client behavior (Explorer, background services) can break the logic. In contrast, Sysmon Event ID 11 (FileCreate) or dedicated File Integrity Monitoring (FIM) solutions provide a deterministic creation signal, which is the industry‑standard for reliable detection. Therefore, the differentiation is not real nor durable using only native audit logs.

Viability

meta/llama-4-maverick-17b-128e-instruct

6.0

Reliably detecting file creation events using native Windows Security logs is challenging due to the similarity in event IDs and AccessMask values between file creation and access.

The task requires analyzing Windows Security logs to detect file creation events, identify the user who created the file, and distinguish between file creation and access. The environment is already set up with a Windows File Server, SIEM integration, and object auditing enabled. The challenge lies in reliably identifying file creation events using native Windows Security logs (4663, 5145, 4656). The proposed approaches, such as filtering 4663 accesses for WriteData/AppendData or correlating multiple event IDs, show promise but are not 100% deterministic. Alternative solutions like using Sysmon or FSRM/File Integrity Monitoring tools could provide more reliable results. A solo or 2-person team can potentially build a v1 solution within 4-12 weeks by focusing on log analysis and correlation, but the complexity of achieving high accuracy and reliability might be underestimated.

Monetization

mistralai/mistral-medium-3.5-128b

7.0

Native Windows logs alone are unreliable for creation tracking, but combining them with Sysmon/FSRM enables a high-margin, scalable SaaS solution.

The idea addresses a real pain point in enterprise environments - reliable file creation tracking in Windows File Servers. The native Windows Security logs (4663, 4656, 5145) are insufficiently deterministic for distinguishing creation from access due to overlapping AccessMask values and event sequences. However, the problem is solvable with complementary tools like Sysmon (Event ID 11) or FSRM, which provide explicit creation events. The venture's value lies in offering a purpose-built solution (e.g., a lightweight agent or SIEM plugin) that correlates native logs with Sysmon/FSRM data, delivering high-confidence creation attribution. Unit economics are favorable: low marginal cost per monitored server, scalable via SaaS pricing ($0.10 - $0.50 per GB monitored/month), and high gross margins (80%+). Channels include direct sales to enterprises and partnerships with SIEM vendors (e.g., Splunk, Microsoft Sentinel). The key risk is competition from built-in tools, but differentiation via accuracy and ease of integration justifies premium pricing.

Market

qwen/qwen3-next-80b-a3b-instruct

7.0

Native Windows audit logs cannot reliably distinguish file creation from access - but enterprises will pay for the best possible approximation when agents like Sysmon are forbidden.

There is a real, paying market for this solution - primarily among enterprise IT security teams, compliance officers, and forensic investigators in regulated industries (finance, healthcare, government) who must track data exfiltration, insider threats, or audit file creation for regulatory requirements like GDPR, HIPAA, or SOX. These organizations already invest in SIEMs and Windows auditing but hit the exact wall described: native Event IDs 4663/5145 are too ambiguous to reliably distinguish creation from access. The unmet need is deterministic file creation attribution without deploying agents. While Sysmon (Event 11) solves this cleanly, many enterprises resist agent deployment due to policy, scalability, or legacy constraints. This creates demand for a pure SIEM-based detection logic that maximizes fidelity from native logs. However, the technical challenge is real: without Sysmon or FSRM, even advanced correlation (e.g., first-seen filename + write flag + 5145 timing) remains probabilistic, not deterministic. This limits adoption to organizations with mature SIEM teams willing to accept ~90% accuracy. The market size is niche but high-value: thousands of enterprises with Windows File Servers and SIEM budgets >$500K/year. The opportunity lies in packaging this as a pre-built SIEM detection pack (KQL/SPL) with tuning guides - not a standalone product. It's not a $100M market, but it's a $10M+ niche with clear buyers who pay for precision in compliance and forensics.

Risk

openai/gpt-oss-120b(fallback #1)

8.0

Dependence on Windows native auditing is a single point of failure vulnerable to regulatory, platform, and budget pressures.

The venture hinges on Windows native audit logs, but three concrete threats can wipe it out in under a year. First, regulators may soon deem such low‑level auditing insufficient for privacy compliance, forcing enterprises to adopt certified third‑party tools and abandoning native solutions. Second, Microsoft could change or retire the SACL auditing APIs in upcoming Server releases, instantly cutting off the data source the product relies on. Third, the primary market - small to midsize businesses - often lacks budget for the required SIEM licensing and advanced audit configurations, leading to rapid churn as they revert to free OS utilities.

Synthesized by meta/llama-3.3-70b-instruct · 7.9s