Technologies

Facebook, Twitter and YouTube face content challenges as Afghanistan falls

From fact-checking to labels, social networks are being put to the test yet again.

A CNN reporter stands in front of a photo of a helicopter flying over the US embassy in Kabul, Afghanistan, a city that has fallen into chaos. Underneath the image, a caption states: «Violent but mostly peaceful transfer of power.»

The image, supposedly a screengrab of the network, circulated widely on Facebook, Twitter and other social media, prompting questions about its authenticity. How could the transfer be considered peaceful, some wondered. Was the language meant to be satire?

Turns out the image was fake.

Reuters and Politifact both fact-checked the image and concluded that it, like so many photos before it, had been digitally altered. The doctored image borrowed a screenshot of CNN correspondent Omar Jimenez from a 2020 broadcast of protests in Kenosha, Wisconsin, over a police shooting. At the time, some conservatives criticized CNN for running the caption «Fiery but mostly peaceful protests after police shooting.»

Altered images and video, such as a doctored version of a Nancy Pelosi speech that made the House Speaker appear drunk, have plagued Facebook and Twitter for years. Now the problem is resurfacing as news pours out of Afghanistan, which quickly fell into turmoil as the US wound down a 20-year war. Just as before, social media outlets are resorting to labels and warnings to caution users about faked content.

On Sunday, Taliban fighters took over Kabul, the capital, and President Ashraf Ghani fled the country. Violence erupted at the city’s international airport, with videos spreading through social media of people clinging to a US military aircraft as it took off and others falling from another plane midair. The Associated Press, citing US senior military officials, reported that at least seven people died at the airport.

The upheaval in Afghanistan poses a set of familiar challenges to social networks, which monitor their platforms for offensive content including graphic imagery. Some Facebook videos of people falling from planes warned users the content didn’t violate its rules but might include violent or graphic content. Similar videos appeared on Twitter and TikTok without a label.

On Facebook and its photo-sharing service Instagram, the doctored CNN image was labeled as altered. «Independent fact-checkers say this information could mislead people,» the label said. The fake CNN caption was also used as a title in a YouTube video with different video footage, and the altered image also spread throughout Twitter, which didn’t add a label. YouTube didn’t label the video and said the video didn’t violate its rules.

Instagram boss: ‘The risk will evolve’

Adam Mosseri, who runs Instagram, told Bloomberg Television that the photo-sharing service bans posts promoting the Taliban, which is covered by its dangerous-organization policies because of US government sanctions.

«We are relying on that policy to proactively take down anything that we can that might be dangerous or that is related to the Taliban in general,» Mosseri said. «Now this situation is evolving rapidly, and with it I’m sure the risk will evolve as well. We are going to have to modify what we do and how we do it to respond to those changing risks as they happen.»

A Facebook spokesman said the company has a dedicated team, «including Afghan nationals and native Dari and Pashto speakers,» to assess the situation in real time.

«Our teams continue to monitor the situation on the ground in Afghanistan, in consultation with our partners, and will take action on any content that violates these policies,» the spokesman said in a statement. Facebook’s online rules prohibit glorifying violence or celebrating the suffering of others but note that it will include a warning screen for some gory content.

Facebook also noted that it bars the Taliban from its services because they’re «sanctioned as a terrorist organization under US law.» The social media giant owns messaging app WhatsApp and reportedly blocked a number being used by the Taliban that’s meant to be a hotline for civilians to report violence, looting and other problems, according to The Financial Times.

From April to June, the social network took action on 7 million pieces of content that contained terrorism, according to Facebook’s Community Standards Enforcement Report released on Wednesday. Facebook didn’t say how much of that content was Taliban-related. The New York Times reported on Wednesday it found more than 100 new accounts and pages on Facebook and Twitter that claim to belong to the Taliban or expressed support for the group.

On YouTube, some news outlets added their own warnings at the beginning of videos that cautioned users the imagery was graphic. But not all did. YouTube added age restrictions and a label to a video of people falling from a plane that was posted by the Hindustan Times, a big Indian newspaper. The label noted the «video may be inappropriate for some users.»

YouTube’s rules don’t allow violent, graphic or shocking content, though they make exceptions for content that is educational, documentary, artistic or scientific. The company said it also surfaces videos from authoritative sources during breaking news events.

In a statement, a YouTube spokesperson said the video-sharing platform would «terminate» accounts it believes are owned and operated by the Taliban because of sanctions and trade compliance laws.

Twitter pointed to its policies against violent organizations and hateful conduct. The company received criticism from some conservatives for allowing Taliban spokesman Zabihullah Mujahid to use its platform. Some activists accused the Taliban of «trying to fish for legitimacy» and pushing out information that conflicts with news reports. The company didn’t immediately answer questions about whether the account violated its rules.

The company has been testing a forum called Birdwatch that lets users flag tweets and write notes with more context. Some of the notes included content about Afghanistan.

In one tweet that was rated as both «not misleading» and «potentially misleading,» Sen. Marco Rubio, a Florida Republican, tweeted that US President Joe Biden «apparently» had «no plans» to speak about Afghanistan. Both of the notes said Rubio tweeted before Biden announced he would be speaking about the topic later on Monday.

Other tweets users rated as misleading note that a video shared by some high-profile conservatives, including US Sen. Ted Cruz, a Texas Republican, «attempt to frame CNN as proponents of the Taliban and their take-over of Afghanistan.» The video shows CNN correspondent Clarissa Ward reporting that Taliban fighters are «just chanting death to America, but they seem friendly at the same time. It’s utterly bizarre.»

TikTok didn’t immediately respond to questions about how it’s moderating content about Afghanistan.

Richard Nieva contributed to this report.

Technologies

Google Is Bringing Gemini AI to Its Smart Home Lineup, Starting Oct. 1

Goodbye, Google Assistant. Hello, Gemini.

It increasingly feels like Google’s AI assistant is omnipresent across our devices and, starting next month, it could also be in your home.

In a post on X on Tuesday, the company teased, «Gemini is coming to Google Home,» and told us to, «Come back October 1.»

At its Made by Google event in August, the company announced Gemini for Home among a slew of other product announcements, so this has been in the works for a while.

Tuesday’s X post teaser appears to show an image of a Nest camera, which Google last upgraded four years ago, suggesting the security camera could be set for a refresh. An upgraded Nest speaker and doorbell, both with 2K camera support, could also be part of the Oct. 1 unveiling.

Google did not immediately respond to CNET’s request for comment.

Android Faithful podcast co-host (and former CNET staffer) Jason Howell is «optimistic» about Gemini replacing Assistant in Google’s smart home products.

«In recent years, I have witnessed my Google Home devices degrading in quality and becoming far less useful for even simple tasks and questions,» Howell tells CNET. «They’ve become buggy and unreliable to the point where I’ve stopped interacting with them for most things.»

Gemini catches dog red-handed

At the Mobile World Congress tech show in Barcelona earlier this year, Howell was impressed by Gemini’s performance with a smart home camera.

«A smart home camera detected a dog that came into the kitchen to steal a cookie off the counter,» Howell recalls. «Through voice interaction, the homeowner could ask the system what happened to the cookie, and, given the video context from the camera and an understanding of what it saw, the system could tell the homeowner that the dog was the culprit.

«This sort of example empowers users to spend less time looking for answers in lieu of simple voice queries that serve them the answer they are looking for with less effort and less time spent.»

Google announced last month that Gemini for Home will eventually replace Google Assistant in its smart home devices. You’ll still activate Gemini with, «Hey Google,» but the advanced AI tech will be able to better interpret more complex and nuanced instructions and questions.

Maybe you’re stumped as to what to make for dinner, so it could be: «Hey Google, what quick pasta dish can I cook in less than an hour?» or, «Give me a recipe for Caesar salad.» Gemini is also designed to work with thermostats and smart lights, so you might tell it to «turn the temp to 68 degrees» and «turn off all the lights except in the kitchen.»

The market for smart home technology is expected to grow by 23% over the next five years, according to Grand View Research.

Technologies

Polar Introduces Loop, a $200 Screenless Wearable

Polar’s first fitness tracker with no screen tracks activity, sleep and overall health, the company says.

Fitness tracking company Polar has launched Loop, a $200 screenless wearable that it says will have no subscription fees. Preorders opened on Sept. 3, and the Polar Loop will start shipping on Sept. 10.

Like other fitness trackers, the Polar Loop will log steps, sleep patterns and daily activity patterns but Polar is touting the lack of a screen as «unobtrusive» and «discreet.» The Loop, which is a wearable band for your wrist, has eight days of battery life with continuous use and stores four weeks’ worth of data. It syncs with the Polar Flow app to view stats and analyze sleep and training data, among other information.

Because it has no buttons, activities can be started in the app or passively with what the company calls «automatic training detection.»

It’s available in the colors Greige Sand, Night Black and Brown Copper. Additional band colors are offered for $20 each.

There’s already a market of no-screen wearables, including the Whoop 5.0 wristband and smart rings such as the Oura Ring 3.

Will the Loop measure up?

Whether the Polar Loop’s attempt at simplifying a fitness wearable works out will largely depend on how well it runs and what it offers compared to other devices.

«The company is clearly tapping into the growing demand for screen-free wearables,» says CNET’s lead writer for wearables, Vanessa Hand Orellana. «It feels like a direct answer to the athlete-favorite Whoop band and even the Oura Ring, both of which collect similar health metrics to display and analyze in their respective apps.»

Hand Orellana says Polar has a good reputation, with its signature heart-rate chest straps, and may win over fans by eschewing the subscription fee that the Oura and Whoop require.

«That said, as with most devices in this space, the real differentiator often comes down to execution… specifically, how well the data translates into clear, actionable insights. Personally, I’m curious to see how the Loop integrates with Polar’s app, which, at least in my experience with their HR straps, hasn’t always been the most intuitive to navigate,» she said.

Technologies

Waymo Is Expanding to Denver and Seattle. Everything to Know About the Robotaxi

The company has also been granted a permit to test its self-driving vehicles in New York City. Here’s everywhere Waymo operates now, and where it’s set to arrive soon.