Testing Grok AI’s Knowledge on Check Fraud
- The way people search for information has changed
- Grok AI's knowledge is tested
- "Check fraud" research is getting easier and easier
"Google it" has become a phrase synonymous with "look it up" in today's society.
And why not? It's easy and convenient to find quick answers to questions you may have. To say the very least, AI has revolutionized online searches by providing personalized answers to your queries.
When we look at check fraud and check fraud detection, some FIs still go to "traditional" search engines like Google to find information about the subject; others explore social media like LinkedIn to see what the experts are saying, and; some reach out to colleagues.
But what does AI have to say?
Testing AI's Knowledge
For our test, we decided to go with Grok AI, as it "excels in real-time analysis" and analyzes inputs from social media (X -- formerly Twitter).
First, we asked a simple question: What is check fraud?
It provided a succinct definition, and cited forged checks, counterfeit checks, check kiting, stolen checks, and washed checks as common variations.
From there we moved on to explore its understanding of previous check fraud technologies, asking Grok; "What is the ASI-16 Check Fraud Detection System?". Grok produced an accurate synopsis, even providing details on how the system is out of date:
The ASI-16 Check Fraud Detection System, developed by Carreker-Antinori, is a legacy analytics tool designed to detect unusual check activity in financial institutions. It uses an archive file to establish account profiles and identify suspicious patterns, such as checks with numbers out of range, previously presented checks, missing checks, accounts with no history, or transactions with dollar amounts significantly higher or lower than the account's average. The system analyzes transactional data, including MICR (Magnetic Ink Character Recognition) line data, check amounts, and serial numbers to flag potential fraud. While effective for its time, ASI-16 is considered outdated (20-30 years old) compared to modern real-time fraud detection systems that incorporate advanced technologies like AI, machine learning, and image analysis to address evolving fraud tactics, such as those seen in mobile remote deposit capture.
For many FIs still utilizing ASI-16, there are clear drawbacks in light of the modern check fraud landscape.
Finally, we decided to see if Grok would be able to tell us how our Anywhere On-Us Fraud systems stacks up against the legacy ASI-16 system:
Upgrading to OrboGraph's Anywhere On-Us Fraud module offers several advanced benefits for financial institutions looking to enhance their check fraud detection capabilities. This module, part of the OrboAnywhere platform, leverages cutting-edge AI and image forensic technologies to address the limitations of legacy systems like the ASI-16.
Grok notes the difference Image Forensic Analyzers provided by our system, including Check Stock Validation (CSV-AI), Automated Signature Verification (ASV-AI), and Alteration Detection (AD-AI).
Grok also notes the following benefits:
- Reduced False Positives and Improved Efficiency:
- Real-Time Fraud Detection Across Channels
- Enhanced Data Integration and Consortium Support
- Comprehensive Fraud Management Tools
- Mitigation of Rising Check Fraud Trends
- Seamless Integration with Other OrboAnywhere Modules:
- Cost and Risk Reduction
- Partnership with Industry Leaders
What Do We Think?
First, we encourage that everyone in the fraud industry exercise due diligence. While it's important to take advantage of advanced search tools on the internet -- via traditional search engines and AI -- speaking to industry colleagues and vendors is also vital in keeping up with the latest check fraud detection innovations and technologies.
As for how Grok performed, it creates narrative answers to prompts vs traditional search engines, which provide links to webpages (which may get mentioned via use of strategic keywords). Grok "harvests" specific details from the gathered sources and amalgamates them into narrative-style text that makes for easy comprehension.
Conclusion: If you're looking for straightforward and narrative-style answers -- with links to sources -- to your check fraud questions, we recommend going with AI chatbots like Grok. Meanwhile, all relevant information about a subject is still available via traditional search engines, though you will need to sift through the results.