April 11 (Reuters) - Instagram will test features that

blur messages containing nudity to safeguard teens and prevent

potential scammers from reaching them, its parent Meta

said on Thursday as it tries to allay concerns over harmful

content on its apps.

The tech giant is under mounting pressure in the United

States and Europe over allegations that its apps were addictive

and have fueled mental health issues among young people.

Meta said the protection feature for Instagram's direct

messages would use on-device machine learning to analyze whether

an image sent through the service contains nudity.

The feature will be turned on by default for users under 18

and Meta will notify adults to encourage them to turn it on.

"Because the images are analyzed on the device itself,

nudity protection will also work in end-to-end encrypted chats,

where Meta won't have access to these images - unless someone

chooses to report them to us," the company said.

Unlike Meta's Messenger and WhatsApp apps, direct messages

on Instagram are not encrypted but the company has said it plans

to roll out encryption for the service.

Meta also said that it was developing technology to help

identify accounts that might be potentially engaging in

sextortion scams and that it was testing new pop-up messages for

users who might have interacted with such accounts.

In January, the social media giant had said it would hide

more content from teens on Facebook and Instagram, adding this

would make it more difficult for them to come across sensitive

content such as suicide, self-harm and eating disorders.

Attorneys general of 33 U.S. states, including California

and New York, sued the company in October, saying it repeatedly

misled the public about the dangers of its platforms.

In Europe, the European Commission has sought information on

how Meta protects children from illegal and harmful content.