- Posted by: aquiliti
- Category: Automation, Business Intelligence, Finance & accounting, Innovation, Procurement
Procurement fraud affects at least 30% of organisations in Australia, and the traditional forms of detection are often ineffective and expensive. Retrospective audits often detect fraudulent activity long after it has been perpetrated. Robotic Process Automation (RPA) technology is increasingly being used to automate the detection of fraud in real time and mitigate economic and reputational risk.
Based on PwC’s Global Economic Crime Survey 2016, at least 52% of organisations (public and private) in Australia have experienced some form of fraud. Of those, 30% have specifically experienced procurement fraud which resulted in economic loss, reputational damage and unnecessary diversion of funds to cover investigation costs.
With 40% of cases reportedly taking 5 years to detect, traditional fraud prevention and detection mechanisms – such as risk management, audits or whistle-blowing – are taking too long to be effective and require significant resources.
With the rise of Robotics Process Automation (RPA) technology, organisations now have an opportunity to automate the early detection of fraud by putting in place checkpoints that could trigger further investigation by humans. Lower entry costs mean that this method of fraud detection is more affordable for more organisations than standard fraud detection software.
RPA applications in practice
Procurement fraud can start at the vendor selection stage and progress right through to the payment of invoices.
If we look at the most recent fraud case investigated by the NSW Independent Commission Against Corruption (ICAC) affecting NSW Department of Finance, Service and Innovation (DFSI) from June to October 2016 (see below for details); an ICT Contractor dishonestly arranged for a company effectively controlled by him to be invoiced on behalf of another legal entity for an amount over the approved budget. An automated workflow fraud detection process could have detected this by checking in real time the approved PO vs. the sum of invoices value, triggering a detailed revision of these transactions and saving DFSI and taxpayers AU$523K.
Below are some common examples of procurement fraud and illustrations of how RPA can be applied for early detection:
|Fraud Type||Description||Example of RPA application|
|Employees conspiring with suppliers||
|Fraudulent supplier set up||
|Avoiding approval thresholds||
|Suppliers’ invoices overstated||
|Supplier GST charge amount||
8 Steps to Automate Fraud Detection
- Identify the priority procurement fraud indicators based on perceived weaknesses in your organisation’s procurement and payment processes.
- Select a pilot fraud indicator to automate. This should be based on the likelihood and consequence of the occurrence and ease of implementation. Checkpoints at the early stages of the procurement process are ideal (i.e. vendor setup, ABN checks).
- Determine data sources, outputs to be reported and timing of reporting (e.g. for severe cases, instant alerts to interested stakeholders)
- Outline detection process for each fraud indicator on which to base the automated flow
- Deploy automated fraud detection process and reporting
- Communicate early outcomes to key stakeholders
- Use results to update current processes and governance to continually reduce organisational risk of fraud repeat occurences
- Assess and include subsequent fraud indicators for automation
Aquiliti uses the latest technologies and innovations to empower possibilities. With input from NSW ICAC, we have designed a set of automated workflows to detect prevalent procurement fraud indicators in organisations. These solutions have been created to proactively and effectively reduce the risk of attacks on your organisation by identifying fraudulent activity in real-time, avoiding fraudulent use of funds and potentially saving you millions of dollars.
Our fraud detection solution complements the objectives of internal audit and your finance functions to reduce the burden of any existing manual detection activities, freeing up your team to focus on delivering your business objectives.
Contact us to understand more about automating fraud detection.