SAN FRANCISCO — Facebook on Friday confronted a new obstacle over what to do with one of its most vital assets — pictures.

The company promised European regulators that it would forgo using facial recognition software to automatically identify its users on the Continent, and delete a database that held millions of pictures uploaded in Europe.

The decision could have wide repercussions on how facial recognition technology — a particularly sensitive technological advance — is used globally as surveillance cameras are increasingly installed in public spaces.

“This is a big deal,’’ said Chris Hoofnagle, a law professor at the University of California, Berkeley, who specializes in online privacy.

The agreement comes as Facebook is under pressure from Wall Street to profit from its vast trove of data, including pictures, and also from regulators worldwide over the use of personal information.

The decision in Europe applies to the ‘‘tag suggestion,’’ a Facebook feature that deploys a sophisticated facial recognition tool to automatically match pictures with names. When a Facebook user uploads a photo of friends, the ‘‘tag suggestion’’ feature can automatically pull up the names of the individuals in the image.

The company quietly and temporarily pulled the plug on ‘‘tag suggestion’’ for all Facebook users several months ago. The company said Friday it was to ‘‘make improvements to the tool’s efficiency’’ and did not say how soon it would be restored.

However, the company promised European regulators Friday that it would reinstate the feature on the Continent only with their approval.

Facebook declined to say under what circumstances the ‘‘tag suggestions’’ would be back online in the United States or elsewhere.

Facebook’s promise to the European regulators is part of an investigation into whether the company’s data collection practices comply with European privacy rules. It was made with regulators in Ireland, where the company has its European headquarters.