Is "Modeling Users for Fraud Detection" too generic for a patent?

Question

In reference to the patent: US20100094767

This is a generic idea that could be easily considered by anyone with a modicum of BI knowledge and AI acumen. Could this patent possibly be granted?

Answer 1

The document you cite is a patent application, not a patent. Applications are frequently rejected. Claims in applications are almost always much broader than what is eventually granted. I will agree that the first claim of the application is almost comically vague and broad. If you go to the US Public Pair, select "Publication Number" and enter 20100094767, you can see the entire application history. This application has, after 9 years of prosecution, been allowed, but the claims look nothing like the application's. Unfortunately the PDF of the allowed claims is an image file so I can't easily quote it here, but take my word it is quite long and specific. Click on the "Image File Wrapper" tab and you can find it. Claim 1 is over two and a half pages long. Long complicated claims are actually weak and narrow. In order to infringe on a claim you need to implement each and every aspect of that claim. Skip even one step and you avoid infringement. The other independent claims are similarly long and complicated.

I didn't bother to read the application so I can't comment on the value or novelty of the invention. I encourage you to read the documents found in the image file wrapper. You can see each of the three non-final rejections, the three final rejections, and each version of the updated claims to try to address the rejections.