Connect with us

Social Networking

Twitter Says Dataminr Monitoring Service Does Not Violate Surveillance Ban

Avatar

Published

on

By Agence France-Presse | Updated: 30 September 2020

Twitter defended letting the service, Dataminr, tap into the flow of public tweets to send alerts to police or other government agencies about plans for protests or civil disobedience, such as those involved in the Black Lives Matter movement.

“Twitter prohibits the use of our developer services for surveillance purposes. Period,” a spokesman for the San Francisco-based company said in reply to an AFP inquiry.

“We see a societal benefit in public Twitter data being used for news alerting, first responder support, and disaster relief.”

The stance provokes a debate as to what exactly constitutes surveillance.

Dataminr is a social media-monitoring service that uses artificial intelligence to comb platforms such as Twitter for user-determined keywords.

In recent months, Dataminr has provided government clients with alerts that include Twitter handles of those posting messages about protest plans or where activists are blocking streets, according to a Wall Street Journal report that cited seeing email copies of alerts.

A Dataminr service called First Alert “notifies first responders about critical events as they’re happening, minimising response time and enabling them to act quickly and confidently,” according to a post at the company’s website.

First Alert relies on public tweets and was built with input from Twitter. Controls were built in to comply with a Twitter policy against surveillance, according to the social media platform.

Twitter said it does not prohibit alerting information about what is happening that can be gleaned from public tweets.

Protests and discussions about the Black Lives Matter movement are major topics on Twitter.

Threat alerts that can keep people out of danger or help support first responders can focus on specific locations, such as parks or schools, and what is happening, according to Twitter.

Twitter said it has audited Dataminr’s suite of products and found no violation of its ban on surveillance.

“We proactively enforce our policies to ensure customers are in compliance and will continue to do so,” the Twitter spokesman said.

“We consistently hold ourselves accountable to rigorous standards, including third-party audits of key products and services like Dataminr.”

Social Networking

Facebook Says Suspected Iranian Hackers Behind US Election Threats Operated in 2019

Avatar

Published

on

By Reuters | Updated: 28 October 2020

Iranian hackers suspected of emailing threatening messages to US voters last week and spreading false information about compromised election systems ran a disinformation campaign last year targeting the Middle East, Facebook said on Tuesday.

US officials blamed Iran last week for thousands of threatening emails and an online video that purported to show hackers breaking into a voter registration system just days before the US presidential election. Tehran has denied the allegations.

Facebook said it had suspended one fake account which attempted to share the video on its site. That account in turn led to more than 20 other accounts on Facebook and Instagram, revealing a dormant disinformation operation that had targeted countries including Israel and Saudi Arabia in 2019, the company said.

Nathaniel Gleicher, Facebook’s head of cybersecurity policy, said the newly-discovered accounts were largely inactive, but had previously attempted to spread claims about an “alleged massacre” at last year’s Eurovision Song Contest in Israel.

US intelligence agencies are still analysing who exactly in Iran commanded the operation and its intent, three people with knowledge of the matter told Reuters last week.

Gleicher said on Tuesday his team found a small number of technical links to a disinformation network suspended in April that was attributed to Iran’s state broadcaster, as well as “connections to individuals associated with the Iranian government.”

Facebook also said it had suspended two pages and 22 Instagram accounts run by people from Mexico and Venezuela that used fake identities and other forms of so-called “coordinated inauthentic behaviour” to post about current events and politics in the United States.

Some of the accounts posed as Americans and posted in Spanish and English about topics including race relations, feminism and the environment, Facebook said. They were identified following a tip from the FBI, it added.

While it was not clear who was behind the activity, some accounts posted captioned pictures previously used by the Internet Research Agency, the Russian organisation US prosecutors have said played a key role in Moscow’s efforts to sway the 2016 US election.

Gleicher said both networks, as well as a third operation targeting Internet users in Myanmar, had been caught before they could attract significant followings.

But he said “malicious actors” were increasingly using concerns about their own election interference attempts to further sow distrust and division.

“We call it perception hacking,” he said. “Rather than actually breaking into a sensitive voter database or using a large social influence campaign, you just play on everyone’s fear that it exists.”

© Thomson Reuters 2020

Continue Reading

Social Networking

Facebook Derails Deception Campaign Ahead of US Presidential Election

Avatar

Published

on

By Agence France-Presse | Updated: 28 October 2020

Facebook on Tuesday said it derailed a fledgling deception campaign aimed at the US that was trying to gain momentum ahead of the presidential election next week.

The effort orchestrated from Mexico posted in English and Spanish on topics including racial injustice, feminism and the environment, using a small bit of content posted in the past by the Russian Internet Research Agency.

Facebook did not link the campaign to Russia, saying it had so far only traced control to unspecified people in Mexico.

The network began creating accounts in April, hiding identities and intent of those involved, according to Facebook head of security integrity Nathaniel Gleicher.

It had only grown to two Facebook Pages and 22 Instagram accounts, according to Facebook.

Those managing accounts or pages used in the “coordinate inauthentic behaviour” campaign claimed to work for what appears to be a fictitious Polish firm.

“Some of these accounts posed as Americans supporting various social and political causes and tried to contact other people to amplify this operation’s content,” Gleicher said.

It was one of three small deception campaigns taken down today at Facebook and Instagram, according to the social network and the latest in a series of efforts by the social network to block efforts to deceptively boost a political candidate or movement.

Each of the networks had few accounts and negligible numbers of followers in what Gleicher said was a sign of Facebook’s success at catching such campaigns quicker.

Catching coordinated deceit efforts faster has triggered a shift in tactics to trying to create a false impression that interference in voting or politics is more pervasive than it actually is, according to Gleicher.

Recent tactics were said to include posing as media outlets or tricking legitimate news agencies into amplifying concerns about social ills or election security.

“We see malicious actors attempt to play on our collective expectation of wide-spread interference to create the perception that they’re more impactful than they in fact are,” Gleicher said.

“We call it perception hacking, an attempt to weaponize uncertainty to sow distrust and division.”

The tactic was used lask week when culprits in Iran spread email messages with unbacked claims of hacking into US voting systems and tried to use Facebook to do the same.

“It’s important that we all stay vigilant, but also see these campaigns for what they are, small and ineffective,” Gleicher said.

One of the campaigns taken down Tuesday was uncovered while digging into the account created to spread the bogus hacking claim.

“This small network originated in Iran and focused primarily on the US and Israel,” Gleicher said.

Continue Reading

Social Networking

Facebook Says Spammers, Scanners Using US Election to Turn Profit Online

Avatar

Published

on

By Reuters | Updated: 22 October 2020

Fraudsters from Albania to Vietnam are posting about US politics and the upcoming presidential election to build fake audiences, maximise clicks and make money online, Facebook said on Wednesday.

In a new report about so-called “inauthentic behaviour” on its platform, Facebook said the November 3 election had become a common lure to trick users into visiting online stores or websites laden with pay-per-view adverts.

“If you are a financially motivated actor who’s trying to make money based on clicks, you are going to use whatever content is going to get you eyeballs,” Nathaniel Gleicher, Facebook’s head of cybersecurity policy, told Reuters.

“And obviously, there’s a lot of attention being paid to what’s happening in the United States around the election.”

After weathering fierce criticism over efforts by political and commercial groups to manipulate its users, Facebook now regularly announces takedowns of online influence operations, such as three networks tied to Russia which it said last month could be used to disrupt the US vote.

Gleicher said Wednesday’s report differed from previous takedown announcements because it dealt with less threatening activity, which was usually financially rather than politically motivated but often confused with foreign interference attempts.

In an interview ahead of the report’s release, he said he wanted to make a distinction between the two types of activity ahead of the US vote and next month’s election in Myanmar, a country where the military and other groups have repeatedly been caught using social media to spread hate and disinformation.

“I want people to be aware of the full range of deception that is happening out there,” Gleicher said. “One of the ongoing challenges is people so often and so regularly mistake a financially motivated scheme to sell T-shirts as an influence operation coming from a foreign government.”
Hot-button issues

There is frequently crossover between the two types of activity, both of which deliberately mislead users with fake accounts and post about “hot-button” issues to build an audience, Facebook said in its report.

But the networks suspended on Wednesday were primarily schemes to amplify content for financial gain, such as by using fake accounts to boost follower numbers or by repeatedly posting spam-like content about popular topics.

Facebook detailed four examples of networks it had suspended between May and September this year, which it said were operated by unconnected groups from countries including Botswana, Bangladesh, Cambodia, Macedonia and the Philippines.

One network operated from Albania and posted about US politics to trick people into following pages that directed them to clickbait websites which generated money through adverts. A page in the network ran under the banner, “We need 1Million Trumpers to Make America Great Again.”

In Myanmar, Facebook said it had suspended 655 pages and 12 groups in August and September that posted about celebrity gossip and local news to attract clicks and views.

“A minority of posts from some of these networks and their ad-heavy websites focused on politics in Myanmar, including support for the military and references to ethnic tensions,” Facebook said. “We did not see evidence of these networks being politically motivated.”

© Thomson Reuters 2020

Continue Reading

Social Networking

Facebook Unveils Machine Learning Translator for 100 Languages

Avatar

Published

on

By Agence France-Presse | Updated: 20 October 2020

Facebook on Monday unveiled software based on machine learning which the company said was the first to be able to translate from any of 100 languages without relying on English.

The open-source artificial intelligence software was created to help the massive social network deliver content better in 160 languages to its more than two billion users around the world.

“This milestone is a culmination of years of Facebook AI’s foundational work in machine translation,” research assistant Angela Fan said in a blog post.

Fan said the new model is more accurate than other systems because it does not rely on English as an intermediary translation step.

“When translating, say, Chinese to French, most English-centric multilingual models train on Chinese to English and English to French, because English training data is the most widely available,” Fan wrote.

“Our model directly trains on Chinese to French data to better preserve meaning. It outperforms English-centric systems by 10 points on the widely used BLEU metric for evaluating machine translations.”

Facebook said it already handles an average of 20 billion translations every day on its news feed, and that it hopes the new system will deliver better results.

“Breaking language barriers through machine translation is one of the most important ways to bring people together, provide authoritative information on COVID-19, and keep them safe from harmful content,” Fan said.

Continue Reading

Social Networking

Twitter Raps Trump COVID-19 Adviser as US Cases Rise

Avatar

Published

on

By Reuters | Updated: 19 October 2020

Twitter on Sunday removed a “misleading” tweet downplaying the efficacy of masks posted by a top coronavirus adviser to the US President Donald Trump, while US cases surged before the November 3 election.

As the Trump administration fends off accusations that its mixed messaging on wearing masks hampered the fight against the coronavirus, Dr. Scott Atlas continued to minimise the importance of masks with a Twitter post on Saturday, saying, “Masks work? NO.”

Twitter removed the tweet on Sunday, saying it violated its misleading information policy on COVID-19, which targets statements that have been confirmed to be false or misleading by subject-matter experts.

The White House had no immediate comment on the decision.

Atlas has downplayed the wearing of masks, a coronavirus containment measure that has been widely endorsed by health experts but not enthusiastically promoted by the president.

Trump, a Republican, is seeking re-election on November 3 against Democratic presidential candidate Joe Biden in the midst of a pandemic that has weakened the economy and killed more than 2,17,000 Americans.

New infections have been rising fast in the United States, according to a Reuters analysis, with more than 69,400 reported on Friday, up from 46,000 a month ago. Total US cases have surpassed 8 million.

Trump, who was hospitalised with the disease for three nights in early October, has been criss-crossing the country in a surge of 11th-hour campaigning as he lags in many public opinion polls. His rallies draw thousands of supporters in close quarters, with many not wearing masks despite federal coronavirus guidelines.

Despite data showing otherwise, Trump has said repeatedly in recent weeks that the country is “rounding the turn” on coronavirus.

On Sunday, Trump again attributed the latest surge in coronavirus cases to more testing, but health experts cite increases in hospitalisations and the rates at which people are testing positive for the virus to show cases are indeed rising.

“The United States shows more CASES than other countries, which the Lamestream Fake News Media pounces on daily, because it TESTS at such a high (and costly) level,” he wrote on Twitter.

Dr. Scott Gottlieb, a former Food and Drug Administration commissioner, said the United States is probably entering the worst phase of the pandemic without a national strategy.

“I think the next three months are going to be very challenging. There’s really no backstop against the spread that we’re seeing,” Gottlieb told CBS’s “Face the Nation” on Sunday.

Hospitalisations were rising in 42 states and there is no intervention short of a vaccine that can thwart the spread, he said. The White House has come out against universal masking, against testing asymptomatic and mildly symptomatic people and wants businesses and schools reopened, he said.

Health and Human Services Secretary Alex Azar urged Americans to continue social distancing, wearing masks and washing hands. “Hang in there with us,” he said on NBC’s “Meet the Press,” adding, “We are so close.”

However, Gottlieb, who sits on the board of vaccine maker Pfizer, said it may not be until February or March until the first tranche of people who are vaccinated are really protected against the virus.

“So it begs the question, ‘what is the strategy?'” he said. “And I think the strategy is just to endure the spread until we get to that vaccine.”

© Thomson Reuters 2020

Continue Reading

Social Networking

Twitter Backtracks, Allows Users to Post Previously Blocked New York Post Article on Joe Biden’s Son

Avatar

Published

on

By Reuters | Updated: 17 October 2020

Twitter on Friday confirmed it reversed its decision to block links to a New York Post article about Democratic presidential candidate Joe Biden’s son, despite reaffirming the ban late on Thursday.

Republicans who had decried Twitter’s earlier actions posted the story freely on the site. “You can now share the bombshell story Big Tech didn’t want you to see,” Arizona Representative Paul Gosar tweeted on Friday morning.

Twitter acknowledged Friday it had stopped blocking links to early versions of the New York Post articles, saying the private information included in them had become widely available in the press and on other platforms.

The company’s policy chief Vijaya Gadde said Thursday night that Twitter had decided to make changes to its hacked materials policy following feedback, but a spokesman told Reuters that the New York Post story would still be blocked for “violating the rules on private personal information.”

“We will no longer remove hacked content unless it is directly shared by hackers or those acting in concert with them,” Gadde said in a series of tweets. “We will label Tweets to provide context instead of blocking links from being shared on Twitter.”

Twitter had initially said the Post story violated its “hacked materials” policy, which bars the distribution of content obtained through hacking, but has provided no details on what materials it viewed as hacked.

Twitter Chief Executive Jack Dorsey said in a tweet Friday morning that “straight blocking of URLs was wrong” and suggested that Twitter instead should have applied tools like labels.

“Our goal is to attempt to add context, and now we have capabilities to do that,” he tweeted.

Tweets of the story successfully published on Friday did not have any labels attached. Twitter declined to answer Reuters questions on whether that was due to an error or a policy decision.

The company had briefly restricted the Twitter account of US President Donald Trump’s re-election campaign after it posted a video that referred to the New York Post story on Thursday.

US Senate Judiciary Committee Chairman Lindsey Graham and Republican Senators Ted Cruz and Josh Hawley said on Thursday the committee would vote on Tuesday on sending a subpoena to Dorsey.

Separately, the Senate Commerce Committee confirmed Friday it will hold an October 28 hearing with Dorsey and the chief executives of Facebook and Google parent Alphabet and will look at “how best to preserve the internet as a forum for open discourse.”

The companies previously confirmed the executives would remotely appear at the hearing.

© Thomson Reuters 2020

Continue Reading

Trending