OTHER
Meta Said to Curtail Election Misinformation Efforts as US Midterm Vote Approaches: Details
Facebook owner Meta is quietly curtailing some of the safeguards designed to thwart voting misinformation or foreign interference in US elections as the November midterm vote approaches.
It’s a sharp departure from the social media giant’s multibillion-dollar efforts to enhance the accuracy of posts about US elections and regain trust from lawmakers and the public after their outrage over learning the company had exploited people’s data and allowed falsehoods to overrun its site during the 2016 campaign.
The pivot is raising alarm about Meta’s priorities and about how some might exploit the world’s most popular social media platforms to spread misleading claims, launch fake accounts and rile up partisan extremists.
“They’re not talking about it,” said former Facebook policy director Katie Harbath, now the CEO of the tech and policy firm Anchor Change. “Best case scenario: They’re still doing a lot behind the scenes. Worst case scenario: They pull back, and we don’t know how that’s going to manifest itself for the midterms on the platforms.”
Since last year, Meta has shut down an examination into how falsehoods are amplified in political ads on Facebook by indefinitely banishing the researchers from the site.
CrowdTangle, the online tool that the company offered to hundreds of newsrooms and researchers so they could identify trending posts and misinformation across Facebook or Instagram, is now inoperable on some days.
Public communication about the company’s response to election misinformation has gone decidedly quiet. Between 2018 and 2020, the company released more than 30 statements that laid out specifics about how it would stifle US election misinformation, prevent foreign adversaries from running ads or posts around the vote and subdue divisive hate speech.
Top executives hosted question and answer sessions with reporters about new policies. CEO Mark Zuckerberg wrote Facebook posts promising to take down false voting information and authored opinion articles calling for more regulations to tackle foreign interference in US elections via social media.
But this year Meta has only released a one-page document outlining plans for the fall elections, even as potential threats to the vote remain clear. Several Republican candidates are pushing false claims about the US election across social media. In addition, Russia and China continue to wage aggressive social media propaganda campaigns aimed at further political divides among American audiences.
Meta says that elections remain a priority and that policies developed in recent years around election misinformation or foreign interference are now hard-wired into company operations.
“With every election, we incorporate what we’ve learned into new processes and have established channels to share information with the government and our industry partners,” Meta spokesman Tom Reynolds said.
He declined to say how many employees would be on the project to protect US elections full time this year.
During the 2018 election cycle, the company offered tours and photos and produced head counts for its election response war room. But The New York Times reported the number of Meta employees working on this year’s election had been cut from 300 to 60, a figure Meta disputes.
Reynolds said Meta will pull hundreds of employees who work across 40 of the company’s other teams to monitor the upcoming vote alongside the election team, with its unspecified number of workers.
The company is continuing many initiatives it developed to limit election misinformation, such as a fact-checking program started in 2016 that enlists the help of news outlets to investigate the veracity of popular falsehoods spreading on Facebook or Instagram. The Associated Press is part of Meta’s fact-checking program.
This month, Meta also rolled out a new feature for political ads that allows the public to search for details about how advertisers target people based on their interests across Facebook and Instagram.
Yet, Meta has stifled other efforts to identify election misinformation on its sites.
It has stopped making improvements to CrowdTangle, a website it offered to newsrooms around the world that provides insights about trending social media posts. Journalists, fact-checkers and researchers used the website to analyse Facebook content, including tracing popular misinformation and who is responsible for it.
That tool is now “dying,” former CrowdTangle CEO Brandon Silverman, who left Meta last year, told the Senate Judiciary Committee this spring.
Silverman told the AP that CrowdTangle had been working on upgrades that would make it easier to search the text of internet memes, which can often be used to spread half-truths and escape the oversight of fact-checkers, for example.
“There’s no real shortage of ways you can organise this data to make it useful for a lot of different parts of the fact-checking community, newsrooms and broader civil society,” Silverman said.
Not everyone at Meta agreed with that transparent approach, Silverman said. The company has not rolled out any new updates or features to CrowdTangle in more than a year, and it has experienced hourslong outages in recent months.
Meta also shut down efforts to investigate how misinformation travels through political ads.
The company indefinitely revoked access to Facebook for a pair of New York University researchers who they said collected unauthorised data from the platform. The move came hours after NYU professor Laura Edelson said she shared plans with the company to investigate the spread of disinformation on the platform around the January 6, 2021, attack on the US Capitol, which is now the subject of a House investigation.
“What we found, when we looked closely, is that their systems were probably dangerous for a lot of their users,” Edelson said.
Privately, former and current Meta employees say exposing those dangers around the American elections have created public and political backlash for the company.
Republicans routinely accuse Facebook of unfairly censoring conservatives, some of whom have been kicked off for breaking the company’s rules. Democrats, meanwhile, regularly complain the tech company hasn’t gone far enough to curb disinformation.
“It’s something that’s so politically fraught, they’re more trying to shy away from it than jump in head first.” said Harbath, the former Facebook policy director. “They just see it as a big old pile of headaches.”
Meanwhile, the possibility of regulation in the US no longer looms over the company, with lawmakers failing to reach any consensus over what oversight the multibillion-dollar company should be subjected to.
Free from that threat, Meta’s leaders have devoted the company’s time, money and resources to a new project in recent months.
Zuckerberg dived into this massive rebranding and reorganisation of Facebook last October, when he changed the company’s name to Meta Platforms. He plans to spend years and billions of dollars evolving his social media platforms into a nascent virtual reality construct called the “metaverse” — sort of like the internet brought to life, rendered in 3D.
His public Facebook page posts now focus on product announcements, hailing artificial intelligence, and photos of him enjoying life. News about election preparedness is announced in company blog posts not written by him.
In one of Zuckerberg’s posts last October, after an ex-Facebook employee leaked internal documents showing how the platform magnifies hate and misinformation, he defended the company. He also reminded his followers that he had pushed Congress to modernise regulations around elections for the digital age.
“I know it’s frustrating to see the good work we do get mischaracterised, especially for those of you who are making important contributions across safety, integrity, research and product,” he wrote on October 5. “But I believe that over the long term if we keep trying to do what’s right and delivering experiences that improve people’s lives, it will be better for our community and our business.”
It was the last time he discussed the Menlo Park, California-based company’s election work in a public Facebook post.
Affiliate links may be automatically generated – see our ethics statement for details.
Catch the latest from the Consumer Electronics Show on Gadgets 360, at our CES 2023 hub.
Meta Platforms used public Facebook and Instagram posts to train parts of its new Meta AI virtual assistant, but excluded private posts shared only with family and friends in an effort to respect consumers’ privacy, the company’s top policy executive told Reuters in an interview.
Meta also did not use private chats on its messaging services as training data for the model and took steps to filter private details from public datasets used for training, said Meta President of Global Affairs Nick Clegg, speaking on the sidelines of the company’s annual Connect conference this week.
“We’ve tried to exclude datasets that have a heavy preponderance of personal information,” Clegg said, adding that the “vast majority” of the data used by Meta for training was publicly available.
He cited LinkedIn as an example of a website whose content Meta deliberately chose not to use because of privacy concerns.
Clegg’s comments come as tech companies including Meta, OpenAI and Alphabet’s Google have been criticized for using information scraped from the internet without permission to train their AI models, which ingest massive amounts of data in order to summarize information and generate imagery.
The companies are weighing how to handle the private or copyrighted materials vacuumed up in that process that their AI systems may reproduce, while facing lawsuits from authors accusing them of infringing copyrights.
Meta AI was the most significant product among the company’s first consumer-facing AI tools unveiled by CEO Mark Zuckerberg on Wednesday at Meta’s annual products conference, Connect. This year’s event was dominated by talk of artificial intelligence, unlike past conferences which focused on augmented and virtual reality.
Meta made the assistant using a custom model based on the powerful Llama 2 large language model that the company released for public commercial use in July, as well as a new model called Emu that generates images in response to text prompts, it said.
The product will be able to generate text, audio and imagery and will have access to real-time information via a partnership with Microsoft’s Bing search engine.
The public Facebook and Instagram posts that were used to train Meta AI included both text and photos, Clegg said.
Those posts were used to train Emu for the image generation elements of the product, while the chat functions were based on Llama 2 with some publicly available and annotated datasets added, a Meta spokesperson told Reuters.
Interactions with Meta AI may also be used to improve the features going forward, the spokesperson said.
Clegg said Meta imposed safety restrictions on what content the Meta AI tool could generate, like a ban on the creation of photo-realistic images of public figures.
On copyrighted materials, Clegg said he was expecting a “fair amount of litigation” over the matter of “whether creative content is covered or not by existing fair use doctrine,” which permits the limited use of protected works for purposes such as commentary, research and parody.
“We think it is, but I strongly suspect that’s going to play out in litigation,” Clegg said.
Some companies with image-generation tools facilitate the reproduction of iconic characters like Mickey Mouse, while others have paid for the materials or deliberately avoided including them in training data.
OpenAI, for instance, signed a six-year deal with content provider Shutterstock this summer to use the company’s image, video and music libraries for training.
Asked whether Meta had taken any such steps to avoid the reproduction of copyrighted imagery, a Meta spokesperson pointed to new terms of service barring users from generating content that violates privacy and intellectual property rights.
© Thomson Reuters 2023
Affiliate links may be automatically generated – see our ethics statement for details.
OTHER
WhatsApp, Instagram and Messenger to Get AI Assistants; Meta Shows Off Image Generation Tool Emu
Meta showcased a host of new products and services, including the Meta Quest 3 mixed reality headset and a pair of smart glasses made in collaboration with Ray-Ban, at its Meta Connect annual conference on Wednesday. Alongside the hardware, the company also announced its own AI assistant, Meta AI, and a variety of AI experiences across Meta’s suite of apps and devices, including AI stickers in Meta apps and AI editing tools for Instagram. Meta AI, a conversational generative AI assistant much like OpenAI’s ChatGPT or Microsoft’s Bing, will be available on WhatsApp, Messenger, and Instagram.
Meta AI will be powered by the company’s custom model that borrows from Meta’s large language model, Llama 2. The AI assistant, Meta said, will provide real-time information in response to text-based queries, trawling the internet via Bing search. Meta AI will also generate images based on text prompts. The AI assistant can help plan hiking trips with your friends in a group chat, prepare recipes, or help with your shopping list. Users can type in “@MetaAI /imagine” inside their chat box and follow it up with descriptive text prompts for what they want the AI assistant to do. Meta AI is also coming to the company’s latest devices, the Meta Quest 3 and the Ray-Ban Meta smart glasses.
In addition to its default AI assistant, the company also showed off AI avatars with distinct personalities. Meta is bringing 28 AI characters, each with a unique backstory and behaviour. These AI characters can be conversed with in WhatsApp, Instagram, and Messenger, and will also include some public figures and influencers that Meta has partnered up with for their likenesses. Famous people coming as AI characters include Dwayne Wade, Kendall Jenner, Mr. Beast, Snoop Dogg and more.
Emu can generate stickers based on user prompts
Photo Credit: Meta
Meta is calling its image generation tool ‘expressive media universe’, or Emu. The tool can also quickly generate AI stickers based on a user’s text prompts inside apps like WhatsApp or Instagram to share with friends. “It’s high-quality, photorealistic,” Meta CEO Mark Zuckerberg said during the presentation. “But, one of the coolest things is the Emu generates that fast. It’s not a minute. It takes five seconds to generate one of these,” he added. This custom sticker generation feature will roll out to select English-language users over the next month in WhatsApp, Messenger, Instagram, and Facebook Stories.
The Facebook parent is also bringing new AI-powered image editing tools, specifically two new features — Restyle and Backdrop — that utilise technology from the Emu tool. Restyle acts as a kind of custom filter that works based on user prompts. Based on single descriptor or a more detail prompt, Restyle will edit your images to reflect a particular mood. And as the name suggests, Backdrop will let users change the background of their images based on custom prompts. Images created using both tools will carry markers that indicate the image is AI-generated. Meta said that Restyle and Backdrop were coming soon to Instagram but did not provide a concrete release date for the same.
Affiliate links may be automatically generated – see our ethics statement for details.
Meta Smart Glasses in collaboration with Ray-Ban were introduced on Wednesday alongside the Meta Quest 3 and other products. Users can livestream videos to Facebook and Instagram using the smart glasses, hands-free. The frame comes with a 12-megapixel camera sensor and an LED unit. This smart wearable succeeds the Ray-Ban stories, which were the company’s first smart glasses and was released in September 2021. However, unlike AR/VR headsets, the Ray-Ban Meta Smart Glasses do not feature a display unit.
Ray-Ban Meta Smart Glasses with standard lenses start at $299 (roughly Rs. 24,999), while Polarized lenses and transition lenses are priced at $329 (roughly Rs. 27,400) and $379 (roughly Rs. 31,500), respectively. It is offered in 150 different custom frame and lens design combinations.
The Ray-Ban Meta Smart glasses are currently available for pre-orders in 15 countries, including the US, Canada, Australia, and European markets. The sale of the smart glasses will start from October 17 in these regions. Meta has not revealed the India launch date of the Ray-Ban Meta Smart Glasses.
Ray-Ban Meta Smart Glasses specifications, features
A 12-megapixel sensor and an LED light, which doubles as a recording indicator are placed within two circular cutouts on either side of the Ray-Ban Meta Smart Glasses frame. Users can use the camera to take photos with a resolution of 3,024 x 4,032 pixels and 1080p videos of up to 60 seconds. With the Meta View app, the users can then share these media files to any other image/ video sharing apps.
The Ray-Ban Meta smart glasses also enable first person perspective livestreaming in which a user can stream whatever they are looking at with the glasses on in real time, to their respective Instagram and Facebook profiles. Users can also use the ‘Hey Meta’ prompt to enable handsfree functions.
Even though the Ray-Ban Meta Smart Glasses do not come with a display, the company claims, that compared to Ray-Ban Stories, the dual open-ear speakers in the new glasses offer less audio leakage and is said to bring up to 50 percent louder sound, deeper bass, and more clarity.
The Ray-Ban Meta Smart Glasses are powered by the new Qualcomm Snapdragon AR1 Gen1 Platform SoC and packs 32GB of inbuilt storage. Claimed to be sleeker in design than its preceding model, Meta says that the glasses come with up to four hours of battery life, and an additional 32 hours with the charging case. One full charge is claimed to take 75 minutes. The glasses also come with an IPX4 rating.
Affiliate links may be automatically generated – see our ethics statement for details.
Enabling developers to create innovative AIs on Messenger and WhatsApp
Meta Used Public Instagram, Facebook Posts to Train Its New AI Assistant
WhatsApp, Instagram and Messenger to Get AI Assistants; Meta Shows Off Image Generation Tool Emu
Meta Smart Glasses in Collaboration With Ray-Ban Launched, Allows Hands-Free Livestreaming
![youtube-shorts-monetization-guide-[how-much-can-you-make?]](https://www.social-ping.com/wp-content/uploads/2023/09/135449-youtube-shorts-monetization-guide-how-much-can-you-make-80x80.png)
YouTube Shorts Monetization Guide [How Much Can You Make?]
WhatsApp Spotted Working on New Colours, Icons for Chat Interface on Android
![top-social-media-manager-interview-questions-and-answers-[2023]](https://www.social-ping.com/wp-content/uploads/2023/09/130190-top-social-media-manager-interview-questions-and-answers-2023-80x80.png)
Top Social Media Manager Interview Questions and Answers [2023]
What Is FedRAMP, and Why Is It So Important?
WhatsApp Android App Could Get New Interface With White Top App Bar, Suggests Beta Update
Solving Espresso’s scalability and performance challenges to support our member base
Introducing Facebook Graph API v18.0 and Marketing API v18.0
Meta to Offer Paid Versions of Facebook and Instagram in Europe to Avoid Ads: Report
-
LINKEDIN1 week agoCareer stories: Influencing engineering growth at LinkedIn
-
OTHER2 weeks agoWhatsApp iPad Support Spotted in Testing on Latest iOS Beta, Improved Group Calls Interface on Android
-
Uncategorized1 week agoVanity Metrics: Definition & Examples for Marketing
-
Uncategorized4 days ago5 B2B Social Media Marketing Tactics That Actually Work
-
OTHER2 weeks agoCCI Said to Have Appointed Former WhatsApp Executive, Government Officials as New Members
-
OTHER1 week agoWhatsApp Passkey Support Reportedly Rolling Out to Beta Testers on Android: How It Works
-
OTHER1 week agoYouTube Announces AI-Enabled Editing Products for Video Creators
-
Uncategorized1 week agoInstagram Emoji Guide: Meanings, Reactions, Ideas