In what could turn into high-profile civil rights class action litigation, an African-American home owner has sued Fannie Mae, charging that its electronic underwriting system discriminates against minority mortgage applicants.

Safiyyah Rahmaan, a 41-year-old homeowner who lives in Wilson, N.C., filed a multi-count suit in U.S. District Court for the District of Columbia. Her complaint represents the first head-on legal attack against Fannie Mae’s “Desktop Underwriter” system. Desktop Underwriter allows brokers and loan officers across the country to submit key applicant data online, and get back a funding and pricing decision from Fannie Mae almost instantaneously.

Rahmaan’s complaint charges that the credit evaluation models and data embedded in the Desktop Underwriter system violate federal fair housing, fair credit and equal credit opportunity statutes. Though filed initially as an individual complaint, some of the law firms representing Rahmaan specialize in consumer and investor class actions against large corporations.

Fannie Mae dismissed the allegations of the suit. Spokeswoman Janice Daue said the corporation’s underwriting system “complies with fair housing laws,” and “does not consider factors such as race or gender in assessing default rates.” To the contrary, Desktop Underwriter “accurately predicts default rates for all borrowers and expands homeownership opportunities,” said Daue.

Rahmaan applied for a $95,000 home mortgage at 6.75 percent last year from Cornerstone Bank in Wilson, N.C., according to the suit. She was rejected for that loan by the bank, and by several additional lenders. The suit charges that these rejections were tied directly to the credit-scoring moels embedded in the Fannie Mae system. Rahmaan eventually obtrained a 10.5 percent loan from another lender.

The suit charges that the “raw credit data upon which (Fannie Mae’s) entire score system is based is undeniably skewed by race.” The complaint cites a 1999 study conducted by mortgage investor Freddie Mac that documented extreme differences in credit ratings among whites, African Americans and Hispanics, at all income levels.

Freddie Mac’s study covered 12,000 Americans’ credit files, and found that while 27 percent of Caucasians had “bad” credit records as defined by researchers, 48 percent of African-Americans and 34 percent of Hispanics fit into that category.

“Fannie Mae’s blithe reliance on such racially-tinged credit reporting data...guarantees divided scores” and unfair treatment of minoity applicants, according to the complaint.

Fannie Mae’s alleged violation of the Fair Credit Reporting Act occurred, according to the suit, in its failure to provide formal notice of any “adverse action,” along with the identity of the credit reporting agency providing the credit information. In effect, said Rahmaan, the credit information-based system run by Fannie Mae rejected her, but gave her no specific information about why.

Rahmaan’s suit asks the court to bar Fannie Mae from “using or relying on racially discriminatory credit scoring systems” and from purchasing loans underwritten by such systems. Consumer groups began weighing into the controversy late last week. Ed Mierzwinski of the Washington D.C.-based Public Interest Research Group said Rahmaan’s complaint “is an important lawsuit that ...has the potential to expose” the inherent problems with Fannie Mae’s automated underwriting. Mierzwinski said his group has done studies of consumer credit files that found one-third of them “contain serious errors that could affect” an applicant’s risk-score in applying for a loan through an automated underwriting system.