विज्ञान

By Sara Merken (Reuters) -Two federal judges admitted in response to an inquiry by U.S. Senate Judiciary Committee Chairman Chuck Grassley that members of their staff used artificial intelligence to help prepare recent court orders that Grassley called "error-ridden." In letters released by Grassley's office on Thursday, U.S. District Judge Henry Wingate in Mississippi and U.S. District Judge Julien Xavier Neals in New Jersey said the decisions in the unrelated cases did not go through their chambers' typical review processes before they were issued. Both judges said they have since adopted measures to improve how rulings are reviewed. Neals, based in Newark, in his letter said a draft decision in a securities lawsuit "was released in error – human error – and withdrawn as soon as it was brought to the attention of my chambers." He said a law school intern used OpenAI's ChatGPT for research without authorization or disclosure. Neals said his chambers has since created a written AI policy and enhanced its review process. Reuters previously reported that research produced using AI was included in the decision, citing a person familiar with the circumstances. Wingate said in his letter that a law clerk in his court in Jackson used Perplexity "as a foundational drafting assistant to synthesize publicly available information on the docket." He said posting the draft decision "was a lapse in human oversight." Wingate had removed and replaced the original order in the civil rights lawsuit and previously declined to give an explanation, saying it contained "clerical errors." The judges did not immediately respond to requests for comment sent to their court staff. Grassley had asked the judges to explain whether AI was used in the decisions after lawyers in the cases said they contained factual inaccuracies and other serious errors. Grassley in a statement on Thursday said he commended the judges for acknowledging the mistakes and urged the judiciary to adopt stronger AI guidelines. "Each federal judge, and the judiciary as an institution, has an obligation to ensure the use of generative AI does not violate litigants' rights or prevent fair treatment under the law," Grassley said. Lawyers have also increasingly faced scrutiny from judges across the country for apparent misuse of AI. Judges have levied fines or other sanctions in dozens of cases over the past few years after lawyers failed to vet the output the technology generated. (Reporting by Sara Merken; Editing by David Bario and Marguerita Choy)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

(Reuters) -Anthropic on Thursday said it would expand the use of Google Cloud's TPU chips, providing the artificial intelligence startup access to the capacity and computing resources required to train the next generations of Claude models. (Reporting by Juby Babu in Mexico City and Zaheer Kachwala in Bengaluru; Editing by Shilpi Majumdar)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

By Arsheeya Bajwa, Max A. Cherney and Stephen Nellis (Reuters) -Intel beat expectations for September-quarter profit as CEO Lip-Bu Tan's drastic cost-cutting measures helped the chipmaker shore up its finances amid a slew of high-profile investments in the company. Shares jumped more than 8% in after-hours trading.   This marks the Santa Clara, California-based company's first earnings announcement after multibillion-dollar investments from Nvidia and Japan's SoftBank as well as an unprecedented U.S. government stake, with investors anticipating a major cash boost.  Intel received the SoftBank money in the third quarter but has not yet received Nvidia's cash, finance chief Dave Zinsner said in an interview with Reuters. The investments are expected to be a major lifeline for Intel, which has been struggling to maintain its dominance in the PC and server central processing unit markets as it competes with Advanced Micro Devices, while repeated attempts to break into the AI chip market commanded by Nvidia have not produced results.  After its share price dropped about 60% last year due to these concerns, Intel stock has risen nearly 90% so far in 2025 thanks to the new investments in the company, helping it outperform the stock of artificial-intelligence chip leader and Wall Street darling Nvidia.  "Shares popped after-hours based on better-than-feared guidance ex-Altera, visible cost and gross margin progress, AI-PC buzz, and $15B of fresh strategic funding that shores the balance sheet," said Michael Schulman, chief investment officer at Running Point Capital.  Nvidia said last month it would invest $5 billion in Intel, giving it a stake of about 4% after new shares are issued. In August, Intel secured $2 billion from SoftBank. After U.S. President Donald Trump called for CEO Tan's resignation over China ties, a hastily arranged Washington meeting produced an unusual deal involving the U.S. government taking a 10% stake for $8.9 billion.  Demand for Intel's chips was strong enough during the third quarter that the company is tight on supply, Zinsner said. One of the factors is that data center operators have realized that they need to upgrade the central processor (CPU) in order to keep pace with the advanced AI chips, he said. Upgrading the CPU gives power and performance advantages that help build and run AI applications. "We're under shipping demand at this point, which I guess is a high-class problem," Zinsner said. For the September quarter, it recorded adjusted gross margins of 40%, beating estimates of 35.7%, while adjusted profit of 23 cents per share also outperformed expectations of a profit of 1 cent per share, according to data compiled by LSEG.  The company will end the year with a workforce that is over one-fifth smaller than last year, it had said in July, while Tan has significantly pared back the expensive manufacturing ambitions of his ousted predecessor, Pat Gelsinger. Tan has aggressively cut costs and divested assets, after Gelsinger's ambitions of turning Intel into a contract manufacturer of chips that would compete against Taiwan's TSMC led to the company in 2024 recording its first annual loss since 1986.  Intel forecast current-quarter revenue between $12.8 billion and $13.8 billion, with a midpoint of $13.3 billion, compared with analysts' average estimate of $13.37 billion. (Reporting by Arsheeya Bajwa in Bengaluru and Max A. Cherney and Stephen Nellis in San Francisco; Additional reporting by Aditya Soni in Bengaluru; Editing by Sayantani Ghosh and Matthew Lewis)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

By Arsheeya Bajwa, Max A. Cherney and Stephen Nellis (Reuters) -Intel beat expectations for September-quarter profit as CEO Lip-Bu Tan's drastic cost-cutting measures helped the chipmaker shore up its finances amid a slew of high-profile investments in the company. Shares jumped more than 8% in after-hours trading.   This marks the Santa Clara, California-based company's first earnings announcement after multibillion-dollar investments from Nvidia and Japan's SoftBank as well as an unprecedented U.S. government stake, with investors anticipating a major cash boost.  Intel received the SoftBank money in the third quarter but has not yet received Nvidia's cash, finance chief Dave Zinsner said in an interview with Reuters. The investments are expected to be a major lifeline for Intel, which has been struggling to maintain its dominance in the PC and server central processing unit markets as it competes with Advanced Micro Devices, while repeated attempts to break into the AI chip market commanded by Nvidia have not produced results.  After its share price dropped about 60% last year due to these concerns, Intel stock has risen nearly 90% so far in 2025 thanks to the new investments in the company, helping it outperform the stock of artificial-intelligence chip leader and Wall Street darling Nvidia.  "Shares popped after-hours based on better-than-feared guidance ex-Altera, visible cost and gross margin progress, AI-PC buzz, and $15B of fresh strategic funding that shores the balance sheet," said Michael Schulman, chief investment officer at Running Point Capital.  Nvidia said last month it would invest $5 billion in Intel, giving it a stake of about 4% after new shares are issued. In August, Intel secured $2 billion from SoftBank. After U.S. President Donald Trump called for CEO Tan's resignation over China ties, a hastily arranged Washington meeting produced an unusual deal involving the U.S. government taking a 10% stake for $8.9 billion.  Demand for Intel's chips was strong enough during the third quarter that the company is tight on supply, Zinsner said. One of the factors is that data center operators have realized that they need to upgrade the central processor (CPU) in order to keep pace with the advanced AI chips, he said. Upgrading the CPU gives power and performance advantages that help build and run AI applications. "We're under shipping demand at this point, which I guess is a high-class problem," Zinsner said. For the September quarter, it recorded adjusted gross margins of 40%, beating estimates of 35.7%, while adjusted profit of 23 cents per share also outperformed expectations of a profit of 1 cent per share, according to data compiled by LSEG.  The company will end the year with a workforce that is over one-fifth smaller than last year, it had said in July, while Tan has significantly pared back the expensive manufacturing ambitions of his ousted predecessor, Pat Gelsinger. Tan has aggressively cut costs and divested assets, after Gelsinger's ambitions of turning Intel into a contract manufacturer of chips that would compete against Taiwan's TSMC led to the company in 2024 recording its first annual loss since 1986.  Intel forecast current-quarter revenue between $12.8 billion and $13.8 billion, with a midpoint of $13.3 billion, compared with analysts' average estimate of $13.37 billion. (Reporting by Arsheeya Bajwa in Bengaluru and Max A. Cherney and Stephen Nellis in San Francisco; Additional reporting by Aditya Soni in Bengaluru; Editing by Sayantani Ghosh and Matthew Lewis)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

By Junko Fujita TOKYO (Reuters) -Japan's Fujikura has become a standout in the red-hot Nikkei share gauge on expectations the century-old wire maker will benefit from increased investment in artificial intelligence data centres that use its optical fibres. The firm got another leg up this week, rising 6% after Sanae Takaichi was elected prime minister following pledges of targeted investment in AI and other key industries. Since the start of 2025, shares in Fujikura are up more than 160%, outpacing the Nikkei's 22% advance to an all-time high. Its market value of about $33 billion now rivals that of better-known Japanese manufacturers such as air-conditioning maker Daikin and construction equipment manufacturer Komatsu. Ups and downs in global equities have been increasingly tied to the fortunes of the so-called "Magnificent Seven" U.S. technology companies at the forefront of AI development, including Nvidia, Amazon, and Microsoft. In Japan, that euphoria has filtered down to major suppliers and investors in the AI sector, such as Advantest, Tokyo Electron and SoftBank Group. Those firms alone account for most of the Nikkei's gain since early September. Fujikura, founded in 1885 during Japan's Meiji Era of modernisation, has been a focus for investors searching through the AI supply chain to find the next batch of winners. "Generative AI is a big theme and key driver for the Japanese market," said Kazuaki Shimada, chief strategist at IwaiCosmo Securities. "Now investors have started looking for new targets that may become the next Fujikura."  Mitsui Kinzoku, which makes materials for data centre servers, is another AI-related high-flyer in the Japanese market, jumping 192% so far in 2025. Another supplier JX Advanced Metals has soared fourfold since its market debut in March. Starting out as a maker of silk and insulated wires, Fujikura went on to produce the world's first optical fibre in 1959. Those fibres are now a key component of AI data centres, with about 75% of the company's output going to overseas customers, such as Google owner Alphabet. A spokesperson said Fujikura had added capacity for fibre production at its existing facility in February, and it announced in August it would spend 45 billion yen ($298.45 million) to add another factory. ($1 = 150.7800 yen) (Reporting by Junko Fujita; Editing by Rocky Swift and Jamie Freed)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

By Junko Fujita TOKYO (Reuters) -Japan's Fujikura has become a standout in the red-hot Nikkei share gauge on expectations the century-old wire maker will benefit from increased investment in artificial intelligence data centres that use its optical fibres. The firm got another leg up this week, rising 6% after Sanae Takaichi was elected prime minister following pledges of targeted investment in AI and other key industries. Since the start of 2025, shares in Fujikura are up more than 160%, outpacing the Nikkei's 22% advance to an all-time high. Its market value of about $33 billion now rivals that of better-known Japanese manufacturers such as air-conditioning maker Daikin and construction equipment manufacturer Komatsu. Ups and downs in global equities have been increasingly tied to the fortunes of the so-called "Magnificent Seven" U.S. technology companies at the forefront of AI development, including Nvidia, Amazon, and Microsoft. In Japan, that euphoria has filtered down to major suppliers and investors in the AI sector, such as Advantest, Tokyo Electron and SoftBank Group. Those firms alone account for most of the Nikkei's gain since early September. Fujikura, founded in 1885 during Japan's Meiji Era of modernisation, has been a focus for investors searching through the AI supply chain to find the next batch of winners. "Generative AI is a big theme and key driver for the Japanese market," said Kazuaki Shimada, chief strategist at IwaiCosmo Securities. "Now investors have started looking for new targets that may become the next Fujikura."  Mitsui Kinzoku, which makes materials for data centre servers, is another AI-related high-flyer in the Japanese market, jumping 192% so far in 2025. Another supplier JX Advanced Metals has soared fourfold since its market debut in March. Starting out as a maker of silk and insulated wires, Fujikura went on to produce the world's first optical fibre in 1959. Those fibres are now a key component of AI data centres, with about 75% of the company's output going to overseas customers, such as Google owner Alphabet. A spokesperson said Fujikura had added capacity for fibre production at its existing facility in February, and it announced in August it would spend 45 billion yen ($298.45 million) to add another factory. ($1 = 150.7800 yen) (Reporting by Junko Fujita; Editing by Rocky Swift and Jamie Freed)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

(Reuters) -Anthropic said on Thursday its Claude models would be trained using up to one million of Google's artificial intelligence chips, worth tens of billions of dollars, as it aims to improve its generative AI offerings in a fast-moving landscape. The deal underscores the immense computing demands of AI firms for training, deployment and continuous inference in GenAI. To secure infrastructure swiftly, developers have been rushing into multi-billion-dollar agreements. Alphabet's Google, also a backer of Anthropic, will provide additional cloud computing services to the startup behind the Claude chatbot. For Google, the deal comes as it is expanding external availability of its in-house tensor processing units, or TPUs, which were historically reserved for internal use. It rents the TPUs through Google Cloud. Anthropic said it chose the TPUs due to their price-performance and efficiency and its existing experience in training and serving its models with the processors. Reuters exclusively reported earlier in October that Anthropic is projecting to more than double and potentially nearly triple its annualized revenue run rate next year, fueled by the rapid adoption of its enterprise products. (Reporting by Juby Babu in Mexico City and Zaheer Kachwala in Bengaluru; Editing by Shilpi Majumdar, Anil D'Silva and Alan Barona)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

RIO DE JANEIRO (Reuters) -U.S. data center operator Equinix is expanding its operations in Brazil, which the firm sees as a priority market, its managing director for Latin America Eduardo Zago said on Thursday. WHY IT'S IMPORTANT Equinix is one of the companies benefiting from the increased demand for data center infrastructure amid a surge in artificial intelligence usage. Brazil has been seeking to become a sustainable infrastructure hub, given the country's abundant supply of renewable energy. The government has also proposed to exempt data center-related equipment from federal taxes to lure investment. BY THE NUMBERS Equinix has eight data centers in Brazil, with a ninth under construction, Zago said during an event in Rio de Janeiro. All are located across the country's southeastern states of Sao Paulo and Rio de Janeiro, he added. Equinix also bought five real estate areas for a potential future expansion in Brazil, according to Zago. However, the firm does not have a plan for them so far, he said, adding that the company will grow depending on demand. KEY QUOTES "Brazil is a priority region for us," Zago said. "It's a market where there is no lack of space nor money," he added, noting the company is in "strong expansion" in the country. CONTEXT In Latin America, Equinix also has data center operations in Mexico, Chile, Peru and Colombia. (Reporting by Rodrigo Viga Gaier in Rio de Janeiro; Writing by Andre Romani; Editing by Aurora Ellis)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

By Renju Jose SYDNEY (Reuters) -A bill to restrict social media for children under 16 will be introduced in the New Zealand parliament, officials said on Thursday, building momentum for parliament's efforts to prevent young people from being harmed while online. The proposed legislation will require social media platforms to conduct an age verification process, similar to Australia's world-first teen social media ban law passed in 2024. A member's bill submitted in May by ruling National Party lawmaker Catherine Wedd to restrict children using social media was selected on Thursday to be introduced in the parliament. The bill has received support from National Party members but its coalition partners have not confirmed whether they will support the bill. Members' bills can be introduced by any lawmaker not in the cabinet and are selected after a ceremonial lottery. It is not immediately clear when the bill will be introduced in the parliament. A New Zealand parliamentary committee has been looking at the impact of social media harm on young people and the roles that government, business, and society should play in addressing those harms. A report is due in early 2026, according to a statement from the committee last week. Prime Minister Christopher Luxon has been raising concerns about harms to mental health from the overuse of social media among young teens, including misinformation, bullying and harmful depictions of body image. Civil-liberties organisation PILLAR said the bill would not protect children online, and instead would create serious privacy risks and restrict online freedom for New Zealanders. "Aligning with international efforts may sound responsible, but it is lazy policymaking," PILLAR Executive Director Nathan Seiuli said in a statement. (Reporting by Renju Jose in Sydney; Editing by Christian Schmollinger)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

(Reuters) -Internet services company VeriSign reported an over 7% rise in third-quarter revenue on Thursday, driven by steady demand for domain registrations as businesses build their online presence.    Businesses are increasingly looking to boost their online presence with easy-to-remember domain names, as users turn to the internet as their first point of contact for their services, benefiting domain registry service providers such as VeriSign. VeriSign acts as the authoritative registry — meaning it is the sole registry — for .com and .net domains, and operates two of the 13 global internet root servers.  The company reported revenue of $419.1 million for the third quarter, ended September 30, a 7.3% increase from the $390.6 million reported in the same period last year. Reston, Virginia-based VeriSign posted a net profit of $213 million, or $2.27 per share, for the quarter, up from $201 million, or $2.07 per share, reported a year ago. VeriSign processed 10.6 million new domain name registrations for .com and .net in the quarter, compared to 9.3 million in the previous year.     (Reporting by Juby Babu in Mexico City; Editing by Alan Barona)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)