Employee Fraud and Internal Corporate Investigations in Kenya

Employee Fraud and Internal Corporate Investigations in Kenya Independent forensic investigations into fraud, collusion and data theft inside your organisation, run so the evidence holds up at the disciplinary hearing, in the Employment and Labour Relations Court and in a criminal case. Instruct Us on Your Matter Call +254 100 177 094 Reviewed by Paul Ooko, Lead Forensic Investigator, Ultimate Forensic Consultants Ltd. Last updated 1 October 2026. Quick answer An internal corporate investigation establishes what an employee or group of employees did, how, and with whose help, using evidence collected lawfully and kept in a verifiable chain of custody. Ultimate Forensic Consultants secures and analyses devices, email, system logs, payment and procurement records and documents, then reports to your board, audit committee, HR or legal team. The report supports dismissal, criminal referral and civil recovery. We are independent of the people under investigation, and we do not decide guilt: that remains the employer’s decision after a fair hearing. The schemes we investigate, and where the evidence sits Internal fraud rarely announces itself. It shows up as small anomalies in records that someone with access can control. Each scheme leaves a different trail. Scheme Typical warning signs Where we look Payroll fraud and ghost employees Staff who cannot be physically verified, duplicate bank accounts, unexplained allowances or overtime Payroll system logs, bank details against HR files, approval trails, ID and document verification Procurement fraud and kickbacks Repeat awards to one supplier, inflated prices, split orders, new vendors with no history Vendor master data, tender files, supplier registration and ownership, communications between buyers and suppliers Fictitious suppliers and false invoices Vendors sharing addresses, phone numbers or bank accounts with staff Link analysis of vendors, staff and bank accounts, invoice metadata and document examination Expense and claims fraud Recycled receipts, altered amounts, claims for unattended events Receipt forensics, email and calendar records, mobile money and card statements Cash skimming and lapping Unreconciled receipts, delayed banking, customer complaints about unrecorded payments Till and ledger data, bank deposits, CCTV, M-Pesa and Paybill reconciliations Stock theft and inventory manipulation Write-offs that repeat, shrinkage that tracks one shift or one storekeeper Stock system audit logs, gate and CCTV records, dispatch documents Data theft and IP leakage Bulk downloads before resignation, files sent to personal email or cloud, a rival launching a copy of your product Forensic images of laptops and phones, email and cloud logs, USB and printing history Conflict of interest and insider-assisted external fraud Undisclosed ownership of a supplier, staff tipping outsiders, repeated losses at one branch Company registry records, communications, system access logs, SIM and mobile money links where relevant Run it properly: five rules that protect your case Most internal fraud cases are weakened by how the employer handled them, not by the fraud. These are the rules we build every engagement around. Secure evidence before anyone is confronted Once a suspect is alerted, files are deleted and phones are reset. We image devices and preserve logs first, quietly. Investigate independently Anyone implicated, or reporting to someone implicated, stays out of the investigation. We report to the board, audit committee or counsel. Collect lawfully and proportionately Access to devices, email and personal data must follow your policies and the Data Protection Act. Unlawfully obtained evidence creates a second problem. Keep the chain of custody Every device and file is logged, hashed and documented, so the evidence can be authenticated under Sections 106A and 106B of the Evidence Act. Give the employee a fair hearing Dismissal that skips notice and a hearing can be found unfair even where the fraud is real. Our report feeds the process. It does not replace it. What UFC delivers Digital evidence acquisition Forensic imaging of laptops, phones, servers and cloud accounts, with recovery of deleted files and messages. Transaction and document analysis Tracing payments and approvals, and forensic examination of invoices, contracts, signatures and altered records. Link and relationship analysis Connecting staff, vendors, bank accounts, phone numbers and companies to expose collusion and undisclosed interests. Structured interviews Planned, documented interviews of witnesses and subjects, conducted without coercion and recorded properly. Asset tracing and recovery support Identifying where diverted funds went, to support civil recovery, freezing applications and settlement negotiations. Report, hearing and court support A clear findings report for the disciplinary panel, a referral pack for the police or EACC, and expert testimony in court. Two illustrative cases: ghost employees and inventory manipulation These scenarios are composites of patterns we see in Kenyan organisations. Names, sectors and figures are changed or indicative, and they are not reports of a specific client’s matter. Payroll fraud: ghost employees on a multi-site payroll What was suspected A manufacturer with several sites noticed that payroll cost was rising faster than headcount. Finance and HR each suspected the other’s records. What we examined Payroll master data and audit logs, HR personnel files, physical verification of staff at each site, bank and mobile money details of every payee, and a forensic image of the payroll officer’s workstation. What the record showed A group of payees could not be verified as real people, with no attendance records and no supervisor who knew them. Several shared the same bank account or phone number. Their records were created and edited after hours under one officer’s login, and approvals came from a supervisor whose credentials were being used by someone else. How the money moved Net pay for the ghost payees was paid into a small set of accounts. Link analysis tied those accounts to a relative of the payroll officer. What the report supported A fair disciplinary hearing, a criminal referral with a complete evidence pack and a civil recovery claim. It also set out the control failures: no segregation of duties, shared credentials and no periodic headcount verification. Inventory manipulation: write-offs that followed one storekeeper What was suspected A distributor saw repeated “damaged goods” write-offs at one warehouse that did not match supplier returns or insurance claims. What we examined Stock system adjustment logs,

Employee Fraud and Internal Corporate Investigations in Kenya Read More »

SIM Swap and M-Pesa Fraud Investigation in Kenya

SIM Swap and M-Pesa Fraud Investigation in Kenya Forensic reconstruction of SIM swaps, fake M-Pesa confirmation messages and mobile money theft, delivered as evidence your advocate can take to court. Instruct Us on Your Matter Call +254 100 177 094 Quick answer M-Pesa fraud investigation establishes how money left an account and who moved it. Ultimate Forensic Consultants examines the phone, SMS records, SIM history and M-Pesa statements to show whether the loss came from a SIM swap, a fake confirmation message or social engineering. We then prepare an expert report supporting admissibility under Sections 106A and 106B of the Evidence Act. We do not reverse transactions: reversals and recovery run through Safaricom, your bank, the police and the courts, and our report is the evidence that drives them. How mobile money is actually stolen in Kenya Most losses fall into one of five patterns. Knowing which one you are dealing with decides what evidence must be preserved first. SIM swap A fraudster obtains a replacement SIM using a forged ID or an insider. Your line goes dead, and they receive your M-Pesa PIN resets and bank OTPs. Fake M-Pesa confirmation message A “payment received” message that never came from M-Pesa, created with SMS spoofing, a message-generator app or an edited screenshot. The seller, landlord or agent releases goods or cash against it. Wrong-number reversal trick You get a real or fake transfer message, then a call: “I sent it by mistake, please return it.” The money you send back is real. The money you “received” never existed or is reversed. Impersonation and PIN capture Callers posing as Safaricom, your bank or a relative talk you into revealing a PIN, an OTP or a reset code. Agent and till fraud A fake deposit message convinces an agent to hand over cash, or a fake till confirmation convinces a shopkeeper to release stock. Businesses lose most to this pattern. The fake M-Pesa message: what the sharp boys show you, and what the record shows A message on a screen proves nothing. It is text, and text can be typed, spoofed or generated. The M-Pesa statement and your actual balance are the record. Here is the gap we look for. What is shown to you [10-character code] Confirmed. Ksh45,000.00 received from JOHN DOE 07XX XXX XXX on 12/9/26 at 2:04 PM. Looks right. Arrives in the usual thread. Often shown in a hurry, with a phone you are not allowed to hold. What the record shows No entry for this code between 2:00 PM and 2:10 PM. Balance unchanged: Ksh12,300.00. The money never arrived. Everything you “saw” was text. What we examine to tell a genuine message from a fake Check Genuine message Fake message Transaction code Unique 10-character code that appears in the M-Pesa statement Missing from the statement, malformed or reused from an older transaction Balance and charges New balance and transaction cost reconcile with the previous balance and tariff Arithmetic, fees or balance do not reconcile Message metadata Stored in the phone’s SMS database with the network’s service-centre details and timestamps in sequence Spoofed or app-generated messages often carry different service-centre data, odd timestamps or sit out of sequence Wording and format Consistent date format, spacing and name format Small differences in spelling, spacing, date style or name case Source Delivered by the network and recoverable from the device Screenshot only, or a message that cannot be traced to the network Simple protection for shops, agents and landlords: never release goods, cash or keys on an SMS or screenshot. Check your balance on the M-Pesa menu or app, or the till or Paybill portal, before you hand anything over. What UFC investigates Phone and SMS forensics Forensic imaging of the handset, recovery of SMS records including deleted messages, and analysis of sender and delivery data. Fake message authentication Opinion on whether a message or screenshot is genuine, spoofed, generated or edited, tested against the M-Pesa statement. M-Pesa statement analysis Reconciling the disputed messages with the statement, identifying receiving numbers and following onward movement of funds. Your advocate then obtains network and bank records by court order. SIM swap reconstruction A timed sequence of the swap: loss of signal, replacement SIM, OTPs, app logins and transfers. This supports claims against a bank or a mobile network. Insider and agent fraud Investigations for banks, mobile network partners, agents and merchants where staff collusion or repeated losses are suspected. Expert report and testimony A court-ready report with the supporting certificate for electronic evidence, and expert testimony where required. The legal framework we work within Computer Misuse and Cybercrimes Act, 2018: computer fraud, computer forgery, identity theft and impersonation, phishing, and interception of electronic messages or money transfers. Penal Code (Cap. 63): obtaining by false pretences, section 313. Evidence Act (Cap. 80), Sections 106A and 106B: admissibility of electronic records and the certificate that must accompany them. Data Protection Act, 2019: personal data compromised in a SIM swap may support a complaint to the Office of the Data Protection Commissioner. National Payment System Act, 2011: the regime governing mobile money providers and their obligations. The first hours after a SIM swap or fake-message loss Stop the loss. Contact your network provider to block the new SIM, and your bank to freeze accounts, cards and mobile banking. Dispute the transaction. Report it to the M-Pesa customer line immediately and forward scam SMS to 333. Keep the phone as it is. Do not delete messages, reset the device or reinstall apps. Screenshots are not a substitute for the original data. Report to the police. File with the DCI Cybercrime Unit or the nearest station and obtain an OB number. Call us for preservation. The earlier the handset is imaged, the more we can recover. Frequently asked questions Can UFC get my stolen M-Pesa money back? We cannot promise recovery and we do not reverse transactions. We establish how the money moved and who is behind it, and our report supports recovery through Safaricom, your

SIM Swap and M-Pesa Fraud Investigation in Kenya Read More »

How to Prove a Digital Document Was Altered: PDF Metadata Examination in Kenya

PDF Metadata Examination Kenya · Digital & Document Forensics The Contract Says 2019. The File Disagrees. Every digital document carries a second, invisible record of its own history — one that isn’t written by whoever typed the visible text. PDF metadata examination in Kenya is how that hidden record gets read, and how backdated contracts and forged agreements get caught. 3Independent metadata sources per document type 106AEvidence Act standard for electronic records 0Alterations made during examination In short A digital document’s visible content and its actual history are recorded separately. PDF properties, Word’s internal revision structure, and a photograph’s EXIF data all capture when a file was truly created, edited, and saved — independent of whatever date appears in the printed or displayed text. PDF metadata examination in Kenya compares these hidden records against the document’s claimed date to establish whether a contract, agreement, or letter is genuine. Why a Digital Document Can’t Fully Hide Its Own History A paper document only shows what’s on the page. A digital one carries a second record most people never think to look at — created automatically by the software that produced it, without the author’s involvement or awareness. This is the foundation of PDF metadata examination in Kenya: comparing what a document claims about itself against what the file format was actually built to record. This matters because the most common way a fabricated document gets caught isn’t a spelling mistake or an inconsistent signature — it’s a mismatch between the story the document tells and the story its own file structure tells. Three sources carry most of that story: PDF properties, Word’s internal revision data, and EXIF metadata in photographed signature pages. 📄 PDF Properties & Metadata Every PDF embeds a Document Information Dictionary and often XMP metadata recording creation software, creation date, and modification history — set by the computer, not the author. 📝 Word Revision History A .docx file is a compressed archive of XML parts. Author names, save timestamps, and sometimes prior revisions can persist inside that structure long after visible track changes are cleared. 📷 EXIF Data Photos of signed signature pages — sent over WhatsApp or email — often carry a capture timestamp, device model, and sometimes GPS coordinates, independent of anything written on the page itself. “You can retype a date. You cannot retype the moment your computer says the file was actually born.” PDF Properties: What the Document Information Dictionary Reveals Open the properties panel of almost any PDF and you’ll find a small set of fields most people ignore: Author, Producer, CreationDate, ModDate, and often the specific software version used to generate it. These fields are written automatically at the moment of export or save — a contract “created” using a version of Adobe Acrobat or Microsoft Word that didn’t exist yet at the claimed contract date is, on its own, close to conclusive. Beyond the visible properties panel, many PDFs also carry XMP metadata — a more detailed, XML-based record embedded in the file that can include edit history, prior filenames, and the specific application and operating system used at each save. Where a PDF has gone through multiple saves or exports, XMP metadata can preserve a trail of those events even after the visible properties panel has been manually edited to show something else. Anatomy of a Backdated PDF What the Document Claims vs. What Its Properties Actually Record A simplified, illustrative property panel — annotated the way an examiner reads it during a PDF metadata examination in Kenya. Document text states:“Executed this 4th day of March, 2019” CreationDate:2026-05-14 11:42:07POSTDATES CLAIM ModDate:2026-05-14 11:47:52SAME-DAY EDIT WINDOW Producer:Microsoft® Word for Microsoft 365VERSION DID NOT EXIST IN 2019 Author:DESKTOP-6XJ2P\\userDEVICE NAME UNRELATED TO SIGNATORY XMP Edit History:2 prior save events detectedCONTRADICTS “ORIGINAL SCAN” CLAIM FAIL Directly contradicts the document’s claimed history WARN Inconsistent, needs corroboration PASS Consistent with genuine timeline Word Revision History: What Survives “Accept All Changes” A .docx file is not a single block of text — it’s a ZIP archive containing multiple XML files that separately describe the document’s content, formatting, comments, and revision data. Clicking “Accept All Changes” removes the visible track-changes markup from the reading view, but it doesn’t always purge every trace of that history from the underlying file structure, especially where a document has been saved rather than freshly exported as a clean copy. Examiners look specifically at the document’s internal author list — every account name that has ever edited the file, in what’s sometimes a longer list than the document’s visible signatories would suggest — alongside embedded save timestamps and, in some cases, remnants of deleted comments or prior paragraph versions. A contract whose internal author history includes a name entirely unconnected to either party, or whose earliest recorded edit postdates the agreement’s claimed signing date, raises exactly the kind of question a court or opposing counsel needs answered. EXIF Data: When the “Original Signed Copy” Is Just a Photo An enormous number of disputed agreements in Kenya don’t arrive as clean PDFs at all — they arrive as a photograph of a signed page, taken on a phone and sent over WhatsApp or email. That photograph carries its own metadata, called EXIF (Exchangeable Image File Format) data, which can include the exact date and time the photo was taken, the camera or phone model used, and — where location services were enabled — GPS coordinates for where the photo was captured. This becomes significant when a photographed “original” signature page is claimed to have been signed at one time and place, but its EXIF capture timestamp — or the compression pattern typical of a specific messaging app’s re-encoding — tells a different story. Even where an image has been forwarded and re-compressed multiple times, forensic examination can often still establish the file lineage and, where original EXIF data survives, the true capture window. Backdated Contracts: Reading the Three Sources Together The strongest findings rarely rest on a single metadata field — they come from

How to Prove a Digital Document Was Altered: PDF Metadata Examination in Kenya Read More »