From X working for restoration in Brazil to police officers barred from social media activity without prior permission in Pakistan, here’s everything we covered at Digital Rights Monitor (DRM) this week.
Read our weekly news roundup.
BRAZIL: X complying with court order for restoration
X has initiated legal proceedings for its restoration in Brazil after the Supreme Court imposed a nationwide ban on the platform for failing to comply with its order, according to a report by Reuters.
The company, formerly known as Twitter, is now complying with the order it initially refused. The Brazilian Supreme Court had directed X to name a new legal representative after the company’s former representative left the country.
The development came on Friday after X named Rachel de Oliveira Conceicao as its new legal representative and submitted her name to the Supreme Court. X has been facing an official investigation into its content moderation practices. The platform, owned by billionaire Elon Musk, was accused of spreading misinformation and hate in April.
On Saturday, the court gave X five days to turn in commercial registries and other documents related to the designation of its new legal representative. Last week, X had confirmed that it was working on resuming its operations in the country after an update to its communications network “inadvertently” made the platform accessible to many users again. Subsequently, the court warned X not to circumvent the ban.
More here
Telegram agrees to share user data with authorities
Telegram has announced it will start sharing data with authorities in case of valid legal requests, marking a significant shift in the platform’s privacy policies.
The announcement came in a Telegram post by the app’s CEO and co-founder Pavel Durov, who has 13 million followers on the platform. The data to be shared includes IP addresses and phone numbers, according to the post.
“We’ve made it clear that the IP addresses and phone numbers of those who violate our rules can be disclosed to relevant authorities in response to valid legal requests,” said Durov, who was arrested last month in France over alleged failure to contain “extremist” content on the encrypted messaging platform.
Durov was released on bail; however, he is not allowed to leave the country during the probe into Telegram. The company has updated its privacy policy on its website as well.
More here
PAKISTAN: Police officers barred from social media activity without approval
The Islamabad police officers and staff have been restricted from expressing personal views online and uploading content to social media platforms without obtaining prior permission, the Associated Press of Pakistan (APP) has reported.
Officers have been barred from expressing personal opinions across social media platforms, publishing articles in print media, and speaking at public events in an official capacity without seeking permission from their respective departments.
The directives were laid out in an office memorandum issued on Tuesday by Inspector General of Police (IG) Syed Ali Nasir Rizvi.
The new rules prohibit officers and staff from taking pictures in their uniform within government premises, official vehicles, or private spaces for personal publicity.
“No officer or official shall upload any kind of secret or official documents, pictures, or content of such documents on social media,” the memorandum reads. Expression of personal, political, or religious opinions online has been barred under the directives.
More here
New tool to detect, fix errors in AI text, says Microsoft
Microsoft has announced the launch of a new feature for its generative artificial intelligence (AI), claiming it will detect inaccuracies in the outputs and fix them.
The “correction” feature, embedded in Microsoft’s Azure AI, will automatically detect and provide revised text, Microsoft says. A preview of the feature is available in the Azure AI Studio, which serves as a suite of safety tools to find errors, detect hallucinations (a wrong answer generated in response to a prompt), and block malicious entries.
The correction feature will be capable of detecting inaccuracies in AI results through their comparison against the source material provided by a user. Microsoft says the new tool will lay out a comprehensive reasoning regarding the detected mistakes, and will help “support the reliability and accuracy” of generative AI application outputs.
“With this enhancement, groundedness detection not only identifies inaccuracies in AI outputs but also corrects them, fostering greater trust in generative AI technologies,” says Microsoft.
More here
Keep following DRM for updates on internet governance, platform takedowns and internet shutdowns, and corporate accountability from around the world
