ChatGPT head of product Nick Turley made the statement while testifying at trial in the case where U.S. Department of Justice seeks to require Google to undertake far-reaching measures restore competition in online search. The judge overseeing the trial in Washington found last year that Google has a monopoly in online search and related advertising.

Google has not offered Chrome for sale. The company plans to appeal the ruling that it holds a monopoly.

‘Better Product’

Turley testified earlier in the day that Google shot down a bid by OpenAI to use its search technology within ChatGPT. OpenAI had reached out to Google after experiencing issues with its own search provider, Turley said, without naming the provider. ChatGPT uses technology from Microsoft’s search engine, Bing.

“We believe having multiple partners, and in particular Google’s API, would enable us to provide a better product to users,” OpenAI told Google, according to an email shown at trial.

OpenAI first reached out in July, and Google declined the request in August, saying it would involve too many competitors, according to the email.

“We have no partnership with Google today,” Turley said.

The DOJ’s proposal to make Google share search data with competitors as one means of restoring competition would help accelerate efforts to improve ChatGPT, Turley said.

Search is a critical part of ChatGPT to provide answers to user queries that are up to date and factual, Turley said. ChatGPT is years away from being able to use its own search technology to answer 80% of queries, he added.

Prosecutors raised concerns in opening statements on Monday that Google’s search monopoly could give it advantages in AI, and that its AI products are another way to lead users to its search engine.

Google has said the case is not about AI, and that it faces robust competition from companies including Meta Platforms and Microsoft.

Ending Exclusive Deals

In August, U.S. District Judge Amit Mehta found that Google protected its search monopoly through exclusive agreements with Samsung Electronics and others to have its search engine installed as the default on new devices.

Google had contemplated deals with Android phone makers such as Samsung that would provide exclusivity for not only its search app, but also for its Gemini AI app and Chrome browser, according to a document shown at trial.

Instead of entering more exclusive agreements, however, Google loosened its most recent deals with device makers Samsung and Motorola and wireless carriers AT&T and Verizon, allowing them to load rival search offerings, other documents showed.

The non-exclusive agreements mirror what Google has said should be the remedy to address Mehta’s ruling. The DOJ wants the judge to go farther, banning Google from making lucrative payments in exchange for installation of its search app.

Google sent letters as recently as last week reiterating that its agreements did not prohibit the companies from installing other AI products on new devices, Google executive Peter Fitzgerald testified on Tuesday.

—

(Reporting by Jody Godoy in Washington, Editing by Nick Zieminski and Deepa Babington)

The post OpenAI Would Buy Google’s Chrome, Exec Testifies at Trial appeared first on GV Wire.

Alberto Carvalho, the district’s superintendent, spoke about Ed in bold terms. In an April speech promoting the software, he promised it would “democratize” and “transform education.” In response to skeptics of AI, he asked, “Why not allow this edutainment approach to capture and captivate their attention, be the motivator?”

One seventh-grade girl who tested the chatbot — personified by a smiling, animated sun — had reported, “I think Ed likes me,” Carvalho said.

Related Story: LA’s Schools Chief Knows What It’s Like to Be Undocumented

A Bold Vision for Ed

Los Angeles agreed to pay a startup company, AllHere, up to $6 million to develop Ed, a small part of the district’s $18 billion annual budget. But just two months after Carvalho’s April presentation at a glittery tech conference, AllHere’s founder and CEO left her role, and the company furloughed most of its staff. AllHere posted on its website that the furloughs were because of “our current financial position.”

AI companies are heavily marketing themselves to schools, which spend tens of billions of dollars annually on technology. But AllHere’s sudden breakdown illustrates some of the risks of investing taxpayer dollars in artificial intelligence, a technology with enormous potential but little track record, especially when it comes to children. There are many complicated issues at play, including privacy of student data and the accuracy of any information offered via chatbots. And AI may also run counter to another growing interest for education leaders and parents — reducing children’s screen time.

Natalie Milman, professor of educational technology at George Washington University, said she often advises schools to take a “wait and see” approach to purchasing new technology. While AI is worthy of use and testing, she said, she warned about schools “talking nebulously about this glorified tool. It has limitations, and we need to ensure we are being critical of what it can do, and its potential for harm and misinformation.”

Related Story: ACLU Sues Defense Department Schools Over Book Bans

Risks and Expert Warnings

AllHere did not respond to interview requests or written questions.

In a statement, Britt Vaughan, a spokesperson for the Los Angeles school district, drew a distinction between distracted students being “consumed by phones during the school day” and students using laptops or tablets to interact with the Ed platform, which he said was “intended to provide individualized educational pathways to address student learning.”

Anthony Aguilar, chief of special education for the district, said that despite the collapse of AllHere, a truncated version of Ed remained accessible to families in the district’s 100 “priority” schools, whose students struggle with academics and attendance.

But that software is not a sophisticated, interactive chatbot. It is a website that gathers information from across many other apps the district uses to track assignments, grades and support services. Students using the site can also complete some learning activities on the platform, such as math problems.

The Ed chatbot promoted by Carvalho was tested with students 14 and older, but it was taken offline to refine how it answers user questions, Aguilar said. The goal is for the chatbot to be available in September, a challenge given that AllHere was supposed to provide ongoing technical support and training to school staff, according to its contract with the district. The district said it hoped AllHere would be acquired and that the new owner would continue services.

Related Story: Money, Not Instruction Time, Is at Heart of Designated Schools Negotiations

Startup Struggles and Shifting Missions

Aguilar said the idea for the software had originated with the district, as part of Carvalho’s plan to help students recover from the academic and emotional effects of the pandemic.

AllHere had won a competitive bidding process to build it, Aguilar said.

But the project represented a vast and unwieldy challenge for the startup, which was best known as a provider of automated text messages from schools to families.

AllHere had attracted $12 million in venture capital funding, according to Crunchbase. Its founder and CEO, Joanna Smith-Griffin, now 33, was featured in Forbes, CBS and other media outlets telling a compelling story. As a former educator whose own students were often absent, she said, she founded AllHere in 2016 to help solve the problem.

Automated text messaging seemed to meet the moment when the COVID-19 pandemic began, and chronic absenteeism became a national crisis. In the spring of 2020, AllHere acquired technology developed by Peter Bergman, an economist and education technology expert. It enabled schools to send “nudges” to parents via text messages about attendance, missing assignments, grades and other issues.

Smith-Griffin often spoke about founding AllHere at the Harvard Innovation Labs, a university program to support student entrepreneurs. According to Matt Segneri, the labs’ executive director, Smith-Griffin’s affiliation with the program occurred while she was an undergraduate and then graduate student at the Harvard Extension School.

Like many small startups, the company shifted its mission over time. Last year, AllHere began talking more about an “AI-powered intuitive chatbot.” AllHere would provide artificial intelligence to schools while also keeping a “human in the loop,” the company said, meaning human moderators would oversee the AI to ensure safety and security — a potentially expensive, labor-intensive proposition.

Stephen Aguilar, a professor of education at the University of Southern California — who is not related to Anthony Aguilar — said it was “a fairly common problem” for ambitious school tech efforts to fail. He formerly worked as a developer of educational software, including some projects that could not be delivered as promised.

“Districts have a lot of complex needs and a lot of safety concerns,” he said. “But they often lack the technical expertise to really vet what they are buying.”

The foray into AI is not the first time Los Angeles has made a big bet on education technology, with questionable returns. Beginning in 2013, under a previous superintendent, the district spent tens of millions of dollars buying iPads preloaded with curriculum materials, but the effort was marred by security concerns and technical mishaps.

In Carvalho’s April speech, at a conference hosted by Arizona State University and GSV Ventures, a venture capital firm, he said the Ed chatbot would have access to student data on test scores, mental health, physical health and family socioeconomic status.

Smith-Griffin joined him onstage to explain that student data would live in “a walled garden” accessible only within “the Ed ecosystem.”

Smith-Griffin did not respond to requests for an interview. Vaughan of Los Angeles schools said the district would protect data privacy and security on the platform “regardless of what happens to AllHere as a company.”

In April, AllHere said it was serving “9,100 schools across 36 states.” According to reporting from The74, an education news site, some of AllHere’s other school district contracts, in the five-figure range, were tiny compared with its deal with Los Angeles, which had already netted the company over $2 million.

Some customers beyond Los Angeles have been told that the company’s services are essentially defunct.

Prince George’s County Public Schools in Maryland learned from AllHere on June 18 that “effective immediately” the startup would no longer be able to provide its text messaging service, a district spokesperson said, because of “unforeseen financial circumstances.”

—

This article originally appeared in The New York Times.

By Dana Goldstein
c. 2024 The New York Times Company

The post AI ‘Friend’ for Public School Students Falls Flat appeared first on GV Wire.

The bipedal robots of various makes and sizes navigated the 21.1-kilometer (13.1-mile) course supported by teams of human navigators, operators, and engineers, in what event organizers say was a first. As a precaution, a divider separated the parallel courses used by the robots and people.

While flesh-and-blood participants followed conventional rules, the 20 teams fielding machines in the Humanoid Robot Half-Marathon competed under tailored guidelines, which included battery swap pit stops.

Robot Victory and Awards

The Sky Project Ultra robot, also known as Tien Kung Ultra, from the Tien Kung Team, claimed victory among the nonhumans, crossing the finish line in 2 hours, 40 minutes and 42 seconds.

Awards were also given out for best endurance, best gait design and most innovative form.

The post Humanoid Robots Run a Chinese Half-Marathon Alongside Human Competitors appeared first on GV Wire.

An AI model “reasons” by breaking down a query into steps and working through them in order. Think of how you learned to solve math word problems in school.

Ideally, to generate citations an AI model would understand the key concepts in a document, generate a ranked list of relevant papers to cite, and provide convincing reasoning for how each suggested paper supports the corresponding text. It would highlight specific connections between the text and the cited research, clarifying why each source matters.

The question is, can today’s models be trusted to make these connections and provide clear reasoning that justifies their source choices? The answer goes beyond citation accuracy to address how useful and accurate large language models are for any information retrieval purpose.

Developing a Benchmark for AI Reasoning

I’m a computer scientist. My colleagues − researchers from the AI Institute at the University of South Carolina, Ohio State University and University of Maryland Baltimore County − and I have developed the Reasons benchmark to test how well large language models can automatically generate research citations and provide understandable reasoning.

We used the benchmark to compare the performance of two popular AI reasoning models, DeepSeek’s R1 and OpenAI’s o1. Though DeepSeek made headlines with its stunning efficiency and cost-effectiveness, the Chinese upstart has a way to go to match OpenAI’s reasoning performance.

Related Story: California Attorney General Declines to Join Musk’s Lawsuit Against OpenAI

The Importance of Sentence-Level Specificity

The accuracy of citations has a lot to do with whether the AI model is reasoning about information at the sentence level rather than paragraph or document level. Paragraph-level and document-level citations can be thought of as throwing a large chunk of information into a large language model and asking it to provide many citations.

In this process, the large language model overgeneralizes and misinterprets individual sentences. The user ends up with citations that explain the whole paragraph or document, not the relatively fine-grained information in the sentence.

Further, reasoning suffers when you ask the large language model to read through an entire document. These models mostly rely on memorizing patterns that they typically are better at finding at the beginning and end of longer texts than in the middle. This makes it difficult for them to fully understand all the important information throughout a long document.

Large language models get confused because paragraphs and documents hold a lot of information, which affects citation generation and the reasoning process. Consequently, reasoning from large language models over paragraphs and documents becomes more like summarizing or paraphrasing.

The Reasons benchmark addresses this weakness by examining large language models’ citation generation and reasoning.

Related Story: Judge Allows Newspaper Copyright Lawsuit Against OpenAI to Proceed

Testing Citations and Reasoning Performance

Following the release of DeepSeek R1 in January 2025, we wanted to examine its accuracy in generating citations and its quality of reasoning and compare it with OpenAI’s o1 model. We created a paragraph that had sentences from different sources, gave the models individual sentences from this paragraph, and asked for citations and reasoning.

To start our test, we developed a small test bed of about 4,100 research articles around four key topics that are related to human brains and computer science: neurons and cognition, human-computer interaction, databases and artificial intelligence. We evaluated the models using two measures: F-1 score, which measures how accurate the provided citation is, and hallucination rate, which measures how sound the model’s reasoning is − that is, how often it produces an inaccurate or misleading response.

Our testing revealed significant performance differences between OpenAI o1 and DeepSeek R1 across different scientific domains. OpenAI’s o1 did well connecting information between different subjects, such as understanding how research on neurons and cognition connects to human-computer interaction and then to concepts in artificial intelligence, while remaining accurate. Its performance metrics consistently outpaced DeepSeek R1’s across all evaluation categories, especially in reducing hallucinations and successfully completing assigned tasks.

OpenAI o1 was better at combining ideas semantically, whereas R1 focused on making sure it generated a response for every attribution task, which in turn increased hallucination during reasoning. OpenAI o1 had a hallucination rate of approximately 35% compared with DeepSeek R1’s rate of nearly 85% in the attribution-based reasoning task.

In terms of accuracy and linguistic competence, OpenAI o1 scored about 0.65 on the F-1 test, which means it was right about 65% of the time when answering questions. It also scored about 0.70 on the BLEU test, which measures how well a language model writes in natural language. These are pretty good scores.

DeepSeek R1 scored lower, with about 0.35 on the F-1 test, meaning it was right about 35% of the time. However, its BLEU score was only about 0.2, which means its writing wasn’t as natural-sounding as OpenAI’s o1. This shows that o1 was better at presenting that information in clear, natural language.

Related Story: OpenAI’s Operator Can Now Control Your Computer by Watching Your Screen

OpenAI Holds the Advantage

On other benchmarks, DeepSeek R1 performs on par with OpenAI o1 on math, coding and scientific reasoning tasks. But the substantial difference on our benchmark suggests that o1 provides more reliable information, while R1 struggles with factual consistency.

Though we included other models in our comprehensive testing, the performance gap between o1 and R1 specifically highlights the current competitive landscape in AI development, with OpenAI’s offering maintaining a significant advantage in reasoning and knowledge integration capabilities.

These results suggest that OpenAI still has a leg up when it comes to source attribution and reasoning, possibly due to the nature and volume of the data it was trained on. The company recently announced its deep research tool, which can create reports with citations, ask follow-up questions and provide reasoning for the generated response.

The jury is still out on the tool’s value for researchers, but the caveat remains for everyone: Double-check all citations an AI gives you.

This article is republished from The Conversation under a Creative Commons license. Read the original article here: https://theconversation.com/popular-ais-head-to-head-openai-beats-deepseek-on-sentence-level-reasoning-249109.

The post Popular AIs Head-to-Head: OpenAI Beats DeepSeek on Sentence-Level Reasoning appeared first on GV Wire.

U.S. District Judge Leonie Brinkema in Alexandria, Virginia, found Google liable for “willfully acquiring and maintaining monopoly power” in markets for publisher ad servers and the market for ad exchanges which sit between buyers and sellers. Publisher ad servers are platforms used by websites to store and manage their ad inventory.

Antitrust enforcers failed to prove a separate claim that the company had a monopoly in advertiser ad networks, she wrote.

Lee-Anne Mulholland, vice president of Regulatory Affairs, said Google will appeal the ruling.

“We won half of this case and we will appeal the other half,” she said, adding that the company disagrees with the decision on its publisher tools. “Publishers have many options and they choose Google because our ad tech tools are simple, affordable and effective.”

Google Shares Drop 2%

Google’s shares were down around 2.1% at midday.

The decision clears the way for another hearing to determine what Google must do to restore competition in those markets, such as sell off parts of its business at another trial that has yet to be scheduled.

The DOJ has said that Google should have to sell off at least its Google Ad Manager, which includes the company’s publisher ad server and ad exchange.

Google now faces the possibility of two U.S. courts ordering it to sell assets or change its business practices. A judge in Washington will hold a trial next week on the DOJ’s request to make Google sell its Chrome browser and take other measures to end its dominance in online search.

Google has previously explored selling off its ad exchange to appease European antitrust regulators, Reuters reported in September.

Brinkema oversaw a three-week trial last year on claims brought by the DOJ and a coalition of states.

Google used classic monopoly-building tactics of eliminating competitors through acquisitions, locking customers in to using its products, and controlling how transactions occurred in the online ad market, prosecutors said at trial.

Google argued the case focused on the past, when the company was still working on making its tools able to connect to competitors’ products. Prosecutors also ignored competition from technology companies including Amazon.com and Comcast as digital ad spending shifted to apps and streaming video, Google’s lawyer said.

—

(Reporting by Jody Godoy in New York; Editing by Chizu Nomiyama and Richard Chang)

The post Google Holds Illegal Monopolies in Ad Tech, US Judge Finds, Allowing US to Seek Breakup appeared first on GV Wire.

These hyper-realistic dolls are nowhere to be found in toy stores, at least for the time being. They are being created using artificial intelligence tools, including ChatGPT.

The AI Doll Trend Takes Off

In recent weeks, social media users have been turning to AI to generate Barbie-fied versions of themselves, their dogs or their favorite famous figures. Anna Wintour has not been spared the AI-doll treatment. Neither has Ludwig van Beethoven.

The trend has frustrated illustrators who oppose the use of unlicensed artwork to train these artificial intelligence tools, and who remain concerned about the effects of AI on their livelihoods. Several have responded by posting similar images of figurines that they illustrated themselves.

Related Story: 5 Easy Steps to Create Your Very Own “Starter Pack” Figurine Meme Image

“HUMAN MADE,” reads a text bubble in the corner of one such illustration by Linh Truong, who depicted herself with her sketchbook and her cat, Kayla.

Truong, 23, an artist who lives in New York City, sees the AI action figures, the latest of several AI portraiture trends, as a way that tech companies are trying to connect with users on a personal level.

“They’re like, ‘We want you to see yourself in our product,’” she said.

Users Embrace AI Self-Portraits

To plenty of people, that’s a tempting possibility.

Suzie Geria, 37, a fitness trainer in Toronto, thought the action figure created for her by ChatGPT was surprisingly realistic. It came with a kettlebell and a cartoon peach to represent the glute-focused class she teaches at a nearby gym.

“It’s kind of cool to see yourself reflected in a cartoon form,” she said. “I think we’re looking at other ways to see ourselves in the world we live in, which is very much online.”

Geria said she had empathy for those who worked in industries that might suffer job loss because of AI. “It’s a tough one, but it’s bringing people joy as well,” she said.

Pat Bassermann, 42, who works in marketing and lives in Andover, Massachusetts, typed a paragraph-long prompt into ChatGPT to create an action figure of himself on Thursday.

“Use this photo of me to create an action figure of myself in a blister pack, in the style like a premium collectible toy,” he wrote, adding requests for grilling tongs and a “relaxed, friendly smile.” He uploaded a headshot, and was presented with an image seconds later.

“Wife & Kids Not Included. Messy House Sold Separately,” reads a line of text at the bottom of the image.

Soon, his three daughters wanted their own versions. In a few more minutes, they were presented with figurines with ponytails, accessorized with ballet slippers, a video game controller and a cup of Boba tea.

Related Story: What to Know About DeepSeek and How It Is Upending Artificial Intelligence

Artists Voice Concerns Over AI

As AI platforms have surged in popularity, their image-generating abilities have come under scrutiny. Artists and musicians have argued that the technology threatens their livelihoods. Deepfake images, many of them explicit, have confounded schools, political campaigns and celebrities.

(The New York Times filed a copyright infringement lawsuit against OpenAI and its partner, Microsoft, accusing them of using published work without permission to train artificial intelligence. They have denied those claims.)

In March, social media was flooded with videos that used ChatGPT to replicate the style of Japanese filmmaker Hayao Miyazaki. In response, some users circulated a clip of Miyazaki calling AI “an insult to life itself” in a 2016 documentary.

Martha Ratcliff, 29, an illustrator in Leeds, England, said she spent years developing a distinctive style of portraiture. She said she felt frustrated every time she saw a new AI portrait trend that ostensibly drew from the work of real artists without compensation.

She gets that it’s fun to hop on a trend, she said. “But I think if you look at the bigger picture, there are a lot of creatives that are worried,” she added. “You just don’t want it to wipe out the whole creative industry.”

She spent about 20 minutes on Saturday making her own hand-drawn rendition of the trend. She depicted herself holding her newborn, surrounded by flowers, colored pencils and a steaming mug that said “mama.”

“A human doing it is so much better than a robot,” she said.

—

This article originally appeared in The New York Times.

By Callie Holtermann/The New York Times via ChatGPT
c. 2025 The New York Times Company

The post AI Action Figures Flood Social Media (Accessories Included) appeared first on GV Wire.

People all over social media have turned themselves into toy figurines with the help of AI image generators.

I’ll help you create your own starter pack figurine using five easy steps:

1. Head over to ChatGPT and create a free account. In order to create AI images, users must sign up first.
2. Enter prompt: The prompt you enter must be very specific and detailed to achieve the desired results. For example: “Generate an action figure-style starter pack image featuring a journalist toy with a press badge, a laptop, and a recording device.”
3. Modify the results but know your limit: Feel free to change the text style, adjust the color or add more items to your figurine. However, since you’re on a free account, ChatGPT limits you on how many changes you can make. As of this posting, the limit was three changes, so choose your adjustments wisely.
4. Download your image: Once you’ve finished, click the download icon on the top right of the image.
5. Post: Submit and post to your favorite social media so the world can see your beautiful creation!

What we started with:

What we ended up with:

This story was inspired by Kapwing’s post.

The post 5 Easy Steps to Create Your Very Own “Starter Pack” Figurine Meme Image appeared first on GV Wire.

Each of the three largest U.S. chip equipment makers – Applied Materials, Lam Research and KLA – may suffer a loss of roughly $350 million over a year related to the tariffs, the sources said. Smaller rivals such as Onto Innovation may also face tens of millions of dollars in extra spending.

The potential billion-dollar cost to the chip equipment industry and talks between industry executives and U.S. officials over several days about those costs are reported here for the first time.

The companies build some of the world’s most highly sought-after chipmaking equipment that can require thousands of specialized parts.

Chip Equipment Makers Lose Billions in Revenue

Chip equipment makers have already lost billions in revenue after former U.S. President Joe Biden implemented a series of export controls aimed at curbing the shipment of advanced semiconductor manufacturing equipment to Chinese entities.

The Trump administration has largely paused the reciprocal tariffs it announced in April. But to spur more U.S. manufacturing, it is weighing further duties on the chip industry and initiated a probe into their imports on Monday.

The estimated costs discussed last week in Washington include lost revenue, primarily for missed sales of less sophisticated equipment to overseas rivals, and the costs of finding and using alternative suppliers for the complex components of chipmaking tools. The estimate also includes tariff compliance costs, such as adding personnel to handle the complexities of following the rules.

Lawmakers and administration officials discussed the tariff costs with chip industry executives and officials from SEMI, an international trade group, as part of an ongoing dialogue.

Applied did not respond to a request for comment. KLA and Lam declined to comment.

The early, rough estimate of $350 million per company could change as the Trump administration’s duties take effect. Quick calculations are hard to make because each chipmaking tool has multiple components, and the ultimate tariff regime is unclear.

The Biden administration cracked down on China’s chip industry over three years to hobble its ability to produce cutting-edge chips used in artificial intelligence, military applications or other ways that could threaten U.S. national security.

The U.S. export controls have spurred China to invest in its domestic chip equipment industry.

—

(Reporting by Max A. Cherney in Toronto; editing by Peter Henderson and Richard Chang)

The post US Tariffs May Cost Chip Equipment Makers More Than $1 Billion, Industry Estimates appeared first on GV Wire.

The acknowledgement appeared to bolster allegations by U.S. antitrust enforcers that Meta had used a “buy or bury” strategy to snap up potential rivals, keep smaller competitors at bay and maintain an illegal monopoly.

It came during Zuckerberg’s second day testifying at the high-stakes trial in Washington, in which the U.S. Federal Trade Commission is seeking to unwind Meta’s acquisitions of prized assets Instagram and WhatsApp.

The case, filed during President Donald Trump’s first term, is widely seen as a test of the new Trump administration’s promises to take on Big Tech companies.

Zuckerberg Believed Instagram Had a Better Camera Product

Asked by an attorney for the FTC whether he thought fast-growing Instagram could be destructive to Meta, then known as Facebook, Zuckerberg said he believed Instagram had a better camera product than Facebook was building.

“We were doing a build vs. buy analysis” while in the process of building a camera app, Zuckerberg said. “I thought that Instagram was better at that, so I thought it was better to buy them.”

Zuckerberg also acknowledged that many of the company’s attempts at building its own apps had failed.

“BUILDING A NEW APP IS HARD”

“Building a new app is hard and many more times than not when we have tried to build a new app it hasn’t gotten a lot of traction,” Zuckerberg told the court.

“We probably tried building dozens of apps over the history of the company and the majority of them don’t go anywhere,” he said.

Zuckerberg’s testimony comes as Meta is defending itself years after the release of damning statements plucked from Facebook’s own documents, like a 2008 email in which he said “it is better to buy than compete.”

The company argues that his past intentions are irrelevant because the FTC has defined the social media market inaccurately and failed to account for stiff competition Meta has faced from ByteDance’s TikTok, Alphabet’s YouTube and Apple’s messaging app.

The FTC accuses Meta of holding a monopoly on platforms used to share content with friends and family, where its main competitors in the United States are Snap’s Snapchat and MeWe, a tiny privacy-focused social media app launched in 2016.

Platforms where users broadcast content to strangers based on shared interests, such as X, TikTok, YouTube and Reddit, are not interchangeable, the FTC argues.

—

(Reporting by Jody Godoy in Washington and Katie Paul in New York; Editing by Alexandra Alper and Mark Porter)

The post At US Antitrust Trial, Meta’s Zuckerberg Admits He Bought Instagram Because It Was ‘Better’ appeared first on GV Wire.

It would also benefit big tech companies like Apple and Samsung and chip makers like Nvidia.

U.S. Customs and Border Protection said items like smartphones, laptops, hard drives, flat-panel monitors and some chips would qualify for the exemption. Machines used to make semiconductors are excluded too. That means they won’t be subject to the current 145% tariffs levied on China or the 10% baseline tariffs elsewhere.

Trump previously said he would consider exempting some companies from tariffs.

Tech Sector Reaction

The move takes off “a huge black cloud overhang for now over the tech sector and the pressure facing U.S. Big Tech,” said Wedbush analyst Dan Ives in a research note.

Neither Apple nor Samsung responded to a request for comment early Saturday. Nvidia declined to comment

The post Trump Administration Says It Will Exclude Some Electronics From ‘Reciprocal’ Tariffs appeared first on GV Wire.