IRS Retreats From Facial Recognition to Verify Taxpayers’ Identities

The IRS has been using the service from a company called ID.me for people to validate and verify their identities before opening online accounts and accessing sensitive personal tax data. The agency was also planning to phase out access to online accounts except through ID.me, starting this summer. ID.me referred questions to the IRS.

The agency didn’t plan to require taxpayers to use the system to pay or file taxes, check their refunds, or check the status of amended returns. But the technology, which includes facial recognition, has been part of an identity-verification system for people to establish online accounts through the IRS website. Those accounts can be used to check on payment plans, look up records and access information about stimulus payments and advanced child tax credits.

Senate Republicans wrote to Mr. Rettig last week to argue that the move threatened personal privacy and created cybersecurity and oversight risks.

“The IRS has unilaterally decided to allow an outside contractor to stand as the gatekeeper between citizens and necessary government services,” the lawmakers, including Sen. Mike Crapo (R., Idaho), wrote.

On Monday, Senate Finance Committee Chairman Ron Wyden (D., Ore.) joined in, asking Mr. Rettig to halt the use of facial-recognition software.

“The government can treat Americans with respect and dignity while protecting against fraud and identity theft,” wrote Mr. Wyden, who applauded Mr. Rettig’s decision later on Monday.

Mr. Wyden also expressed concerns about racial and gender bias in facial-recognition systems.

“Many facial recognition technologies are biased in ways that negatively impact vulnerable groups, including people of color, women, and seniors,” he wrote.

Entrusted with millions of Americans’ financial information, the IRS emphasizes data security and taxpayer privacy, but it has sometimes struggled to meet those goals.

Nearly a decade ago, the IRS struggled to combat tax-refund fraud combined with identity theft. In those cases, identity thieves would use information stolen from outside the IRS—like from employers or healthcare providers—to file fraudulent tax returns and claim refunds before the legitimate taxpayers did. Enforcement and changes to automated systems that flagged questionable tax returns helped stem those schemes.

And last year, federal authorities said they would investigate after the news organization ProPublica released the tax information of several wealthy Americans. The potential breach raised questions about whether there were gaps in IRS data systems. The agency hasn’t provided any detail on what happened and investigations have yet to yield prosecutions or any publicly available reports.

Write to Richard Rubin at richard.rubin@wsj.com and Laura Saunders at laura.saunders@wsj.com