New research reveals that sorbitol, a common sugar alcohol used in “low-calorie” foods, can behave much like fructose once inside the body, potentially contributing to liver strain and metabolic dysfunction. Sweeteners such as aspartame, found in Equal packets, sucralose (Splenda), and various sugar alcohols are often viewed as better choices than foods containing refined sugar […]
Source link
Cloudflare is a platform that provides services including security and traffic management and optimisation for websites and applications. File.
| Photo Credit: AP
Multiple online platforms including U.S, President Donald Trump’s Truth Social network were briefly down Friday (December 5, 2025), with American internet services company Cloudflare acknowledging its systems were to blame.
An error message mentioning Cloudflare was displayed to people trying to visit Truth Social and other sites, such as graphic design platform Canva, until around 0930 GMT.
Also read | How much of the Internet actually runs on Cloudflare
“We are aware of the issue impacting the availability of Cloudflare’s network,” the company’s chief technology officer Dane Knecht posted on the X social network at 0920 GMT, adding that the cause was “not an attack”.
Many X users had complained of problems connecting to other sites earlier in the day.
Cloudflare is a platform that provides services including security and traffic management and optimisation for websites and applications.
It claims to manage around 20% of global internet traffic.
A recent outage affecting its services in mid-November took many widely-used sites offline for several hours.
Published – December 05, 2025 04:52 pm IST
An oversight board created by Facebook to review content-moderation decisions trumpeted improved transparency and respect for people’s rights in a survey of their first five years of work on Thursday, while acknowledging “frustrations” to their arm’s-length role.
Facebook, since renamed Meta, announced the Oversight Board, often referred to as the group’s “supreme court”, at a 2018 nadir of public trust in the tech giant.
The Instagram and WhatsApp owner’s image had been tarnished by episodes like the Cambridge Analytica data-breach scandal and dis- and misinformation around crucial public votes such as Brexit and the 2016 U.S. presidential election.
The Oversight Board began its work in 2020, staffed with prominent academics, media veterans and civil society figures.
It reviews selected cases where people have appealed against Meta’s moderation decisions, issuing binding rulings on whether the company was right to remove content or leave it in place.
It also issues non-binding recommendations on how lessons from those cases should be applied to updating the rules for billions of users on Meta’s Facebook, Instagram and Threads platforms.
Over the past five years, the board has secured “more transparency, accountability, open exchange and respect for free expression and other human rights on Meta’s platforms”, it said in a report.
The board added that Meta’s oversight model, unusual among major social networks, could be “a framework for other platforms to follow”.
The board, which is funded by Meta, has legal commitments that bosses will implement its decisions on individual pieces of content.
But the company is free to disregard its broader recommendations on moderation policy.
“Over the last five years, we have had frustrations and moments when hoped-for impact did not materialize,” the board wrote.
Some outside observers of the tech giant are more critical.
“If you look at the way that content moderation has changed on Meta platforms since the establishment of the board, it’s rather gotten worse,” said Jan Penfrat of Brussels-based campaigning organisation European Digital Rights (EDRi).
Today on Facebook or Instagram, “there is less moderation happening, all under the guise of the protection of free speech,” he added.
Effective oversight of moderation for hundreds of millions of users “would have to be a lot bigger and a lot faster”, with “the power to actually make systemic changes to the way Meta’s platforms work”, Penfrat said.
One major outstanding issue is chief executive Mark Zuckerberg’s surprise decision in January to axe Meta’s US fact-checking programme.
That scheme had employed third-party fact checkers, AFP among them, to expose misinformation disseminated on the platform.
In April, the Oversight Board said the decision to replace it with a system based on user-generated fact-checks had been made “hastily”.
Its recommendation from that time for “continuous assessments of the effectiveness” of the new system is currently marked as “in progress” on the company’s website.
Last month, the Oversight Board said it would fulfil Meta’s request for its advice on expanding worldwide the so-called “Community Notes” programme.
The company said it needed help “establishing fundamental guiding principles” for rolling out the scheme and identifying countries where it might not be appropriate, for example due to limits on freedom of expression.
Looking ahead, “the Board will be widening its focus to consider in greater detail the responsible deployment of AI tools and products,” the report said.
Zuckerberg has talked up plans for deeper integration of generative artificial intelligence into Meta’s products, calling it a potential palliative to Western societies’ loneliness epidemic.
But 2025 has also seen mounting concern over the technology, including a spate of stories of people killing themselves after extended conversations with AI chatbots.
Many such “newly emerging harms… mirror harms the Board has addressed in the context of social media”, the Oversight Board said, adding that it would work towards “a way forward with a global, user rights-based perspective”.
Published – December 05, 2025 09:33 am IST
Speaking at the Reuters NEXT conference in New York, Robby Stein, vice president of product for Google Search, addressed concerns [File]
| Photo Credit: REUTERS
A Google search executive on Thursday pushed back against fears that its AI-powered search will harm web publishers and its advertising business, calling the technology an “expansionary moment” for the internet.
Speaking at the Reuters NEXT conference in New York, Robby Stein, vice president of product for Google Search, addressed concerns that the company’s efforts to provide more AI-powered features — which deliver direct answers to queries — would reduce traffic to external websites.
“Google sends billions and billions and billions of clicks out every single day, and the outbound clicks are largely stable. So that’s actually not changing,” Stein said, adding that new ways people are searching, such as using their phone cameras or asking long, complex questions, are creating new opportunities. “We think over time that’s expansionary. The pie is growing very, very fast.”
Stein also sought to allay investor concerns that a shift to conversational AI could disrupt Google’s lucrative advertising model. He compared the current transition to the migration from desktop to mobile, saying advertising will evolve to fit the new format.
Ads in an AI chat experience can be “incredibly helpful,” he said, citing an example of a user with a raccoon problem being offered relevant products. Stein’s comments came as reports emerged that rival OpenAI had declared a “code red” to improve ChatGPT amid growing competition from Google. The tech giant Alphabet has appeared to wave off early threats to its core business with more diverse experiences, frontier model launches and viral products such as its photo-generation tool Nano Banana. Shares of Google have risen nearly 67% this year, fueled by growth in its cloud unit.
Stein’s remarks counter a growing narrative from media organisations that AI will decimate their traffic and revenue, while also reassuring Wall Street that Google’s core cash-generating machine is not at risk as it revamps its flagship product.
“Google cares about the web more than anyone,” Stein said, describing AI as a “powerful discovery engine” that will help users go deeper by linking out to sources. When asked whether Google felt “unshackled” to innovate following a U.S. antitrust ruling that allows Google to keep its Chrome browser, Stein downplayed the case’s impact. “The model’s capability is the thing that’s driving the innovation and the excitement,” he said.
Google’s push into generative AI for its search engine, which followed a clumsy start in 2023, has met with alarm from the publishing industry. Several media leaders have warned of a sharp decline in referral traffic, a fear backed by a Pew Research study this year that found AI summaries made people less likely to click on source links.
Published – December 05, 2025 10:27 am IST
The United States is poised to lose its measles-free status next year. If that happens, the country will enter an era in which outbreaks are common again.
More children would be hospitalized because of this preventable disease. Some would lose their hearing. Some would die. Measles is also expensive. A new study — not yet published in a scientific journal — estimates that the public health response to outbreaks with only a couple of cases costs about $244,000. When a patient requires hospital care, costs average $58,600 per case. The study’s estimates suggest that an outbreak the size of the one in West Texas earlier this year, with 762 cases and 99 hospitalizations, costs about $12.6 million.
America’s status hinges on whether the country’s main outbreaks this year stemmed from the big one in West Texas that officially began Jan. 20. If these outbreaks are linked, and go on through Jan. 20 of next year, the U.S. will no longer be among nations that have banished the disease.
“A lot of people worked very hard for a very long time to achieve elimination — years of figuring out how to make vaccines available, get good vaccine coverage, and have a rapid response to outbreaks to limit their spread,” said Paul Rota, a microbiologist who recently retired from a nearly 40-year career at the Centers for Disease Control and Prevention.
Instead of acting fast to prevent a measles comeback, Robert F. Kennedy Jr., a lawyer who founded an anti-vaccine organization before taking the helm at the Department of Health and Human Services, has undermined the ability of public health officials to prevent and contain outbreaks by eroding trust in vaccines. The measles vaccine is safe and effective: Only 4% of nearly 1,800 confirmed U.S. cases of measles this year have been in people who had received two doses.
Kennedy has fired experts on the vaccine advisory committee to the CDC and has said, without evidence, that vaccines may cause autism, brain swelling, and death. On Nov. 19, scientific information on a CDC webpage about vaccines and autism was replaced with false claims. Kennedy told The New York Times that he ordered the change.
“Do we want to go back into a prevaccine era where 500 kids die of measles each year?” asked Demetre Daskalakis, a former director of the CDC’s national immunization center, who resigned in protest of Kennedy’s actions in August. He and other scientists said the Trump administration appears to be occupied more with downplaying the resurgence of measles than with curbing the disease.
HHS spokesperson Andrew Nixon said in a statement that vaccination remains the most effective tool for preventing measles and that the “CDC and state and local health agencies continue to work together to assess transmission patterns and ensure an effective public health response.”
Looking for Links
CDC scientists are indeed tracking measles, alongside researchers at health departments and universities. To learn whether outbreaks are linked, they’re looking at the genomes of measles viruses, which contain all their genetic information. Genomic analyses could help reveal the origin of outbreaks and their true size, and alert officials to undetected spread.
Scientists have conducted genomic analyses of HIV, the flu, and covid for years, but it’s new for measles because the virus hasn’t been much of a problem in the U.S. for decades, said Samuel Scarpino, a public health specialist at Northeastern University in Boston. “It’s important to get a surveillance network into place so that we could scale up rapidly if and when we need it,” he said.
“We are working with the CDC and other states to determine whether what we’re seeing is one large outbreak with continued spread from state to state,” said Kelly Oakeson, a genomics researcher at the Utah Department of Health and Human Services.
At first glance, the ongoing outbreak in Utah and Arizona, with 258 cases as of Dec. 1, seems linked to the one in Texas because they’re caused by the same strain of measles, D8-9171. But this strain is also spreading throughout Canada and Mexico, which means the outbreaks could have been sparked separately from people infected abroad. If that happened, this technicality could spare the U.S. from losing its status, Rota said. Being measles-free means the virus isn’t circulating in a country continuously year-round.
Canada lost its measles-elimination status in November because authorities couldn’t prove that various outbreaks from the D8-9171 strain were unrelated, said Daniel Salas, executive manager of the comprehensive immunization program at the Pan American Health Organization. The group, which works with the World Health Organization, includes health officials from countries in North, South, and Central America, and the Caribbean. It makes a call on measles elimination based on reports from scientists in the countries it represents.
Early next year, PAHO will hear from U.S. scientists. If their analyses suggest that measles has spread continuously for a year within the U.S., the organization’s director may revoke the country’s status as measles-free.
“We expect countries to be transparent about the information they have,” Salas said. “We will ask questions, like, ‘How did you determine your findings, and did you consider other angles?’”
In anticipation of PAHO’s assessment, Oakeson and other researchers are studying how closely the D8-9171 strains in Utah match others. Instead of looking at only a short snippet of genes that mark the strain, they’re analyzing the entire genome of the measles virus, about 16,000 genetic letters long. Genetic mutations occur naturally over time, and the accumulation of small changes can act like a clock, revealing how much time has ticked by between outbreaks. “This tells us the evolutionary history of samples,” Oakeson said.
For example, if one child directly infects another, the kids will have matching measles viruses. But measles viruses infecting people at the start of a large outbreak would be slightly different than those infecting people months later.
Although the Texas and Utah outbreaks are caused by the same strain, Oakeson said, “more fine-grained details are leading us to believe they aren’t super closely related.” To learn just how different they are from each other, scientists are comparing them with measles virus genomes from other states and countries.
Ideally scientists could pair genetic studies with shoe-leather investigations into how each outbreak started. However, many investigations have come up dry because the first people infected haven’t sought care or contacted health departments. As in West Texas, the outbreak in Utah and Arizona is concentrated in close-knit, undervaccinated communities that are leery of government authorities and mainstream medicine.
Researchers are also trying to learn how many measles cases have gone undetected. “Confirmed cases require testing, and in some communities, there’s a cost to going to the hospital to get tested: a tank of gas, finding a babysitter, missing work,” Andrew Pavia, an infectious disease doctor at the University of Utah, said. “If your kid has a measles rash but isn’t very sick, why would you bother?”
Subtle Surveillance
Pavia is part of a nationwide outbreak surveillance network led by the CDC. A straightforward way to figure out how large an outbreak is would be through surveys, but that’s complicated when communities don’t trust public health workers.
“In a collaborative setting, we could administer questionnaires asking if anyone in a household had a rash and other measles symptoms,” Pavia said, “but the same issues that make it difficult to get people to quarantine and vaccinate make this hard.”
Instead, Pavia and other researchers are analyzing genomes. A lot of variation suggests an outbreak spread for weeks or months before it was detected, infecting many more people than known.
A less intrusive mode of surveillance is through wastewater. This year, the CDC and state health departments have launched efforts to test sewage from households and buildings for measles viruses that infected people shed. A study in Texas found that this could function as an early warning system, alerting public health authorities to an outbreak before people show up in hospitals.
The quiet research of CDC scientists stands in stark contrast to its dearth of public-facing actions. The CDC hasn’t held a single press briefing on measles since President Donald Trump took office, and its last publication on measles in the agency’s Morbidity and Mortality Weekly Report was in April.
Rather than act fast to limit the size of the Texas outbreak, the Trump administration impeded the CDC’s ability to communicate quickly with Texas officials and slowed the release of federal emergency funds, according to investigations by KFF Health News. Meanwhile Kennedy broadcast mixed messages on vaccines and touted unproven treatments.
Daskalakis said that as the outbreak in Texas worsened, his CDC team was met by silence when they asked to brief Kennedy and other HHS officials.
“Objectively they weren’t helping with the Texas outbreak, so if we lose elimination, maybe they’ll say, ‘Who cares,’” Daskalakis said.
Nixon, the HHS spokesperson, said Kennedy responded strongly to the Texas outbreak by directing the CDC to help provide measles vaccines and medications to communities, expediting measles testing, and advising doctors and health officials. The U.S. retains its elimination status because there’s no evidence of continuous transmission for 12 months, he added.
“Preliminary genomic analysis suggests the Utah and Arizona cases are not directly linked to Texas,” the CDC’s acting director, Deputy HHS Secretary Jim O’Neill, wrote on the social platform X.
Given Kennedy’s distortions of data on vitamin A, Tylenol, and autism, Daskalakis said the Trump administration may insist that outbreaks aren’t linked or that PAHO is wrong.
“It will be quite a stain on the Kennedy regime if he is the health secretary in the year we lose elimination status,” he said. “I think they will do everything they can to cast doubt on the scientific findings, even if it means throwing scientists under the bus.”
MP Sudha Murty speaks in the Rajya Sabha during the Winter session of Parliament, in New Delhi
| Photo Credit: Sansad TV
Rajya Sabha Nominated MP Sudha Murty on Friday (December 5, 2025) urged the Government to set norms on the portrayal of children on social media platforms, as it is essential to inculcate a good value system among the younger generation.
Raising the issue during the Zero Hour, she expressed serious concern about how children are being depicted on social media, and cited regulations on the portrayal of children from several advanced countries like France.
Parliament Winter session LIVE – December 5, 2025
“Children are our future, and we should develop our children in a good value system, good education, sports and many other activities,” Ms. Murty said.
She stated that social media has now become very popular and there are many advantages, but there are also flip sides to it.
Many parents expose their innocent children on social media platforms, she said.
“…they are exposed to different types of costumes, dress, commercialisation, and put them on social media so that they can have 10,000, 1 million and half a million followers. This will help them (parents) financially a lot, I am aware of it, but what happens to the child?” she asked.
Ms. Murty said the children become a source of income to parents, but it affects the child’s psychology in the long term.
“The child will lose innocence. The child does not give permission because the child is not aware of it. It will affect the children’s psychology. They will not learn how to do any social activity, or sports, or even receive a good education,” she observed.
Ms. Murty requested the Central government to regulate this.
She noted that the government has done fantastic work in regulating the portrayal of children in advertisements and the film industry.
“You are regulating in the advertisement, children’s advertisement, children working or acting in a film. All those things have been taken care of, and there are strict laws implemented. Whereas when it comes to social media, it is not done,” Murty said.
In the absence of regulations and restrictions, she warned that this would cause a great problem in future for our children.
“There should be a restriction. Children cannot use a certain kind of dress and dance…because this is not the way how we can bring up the next generation of our children,” Ms. Murty added.
Published – December 05, 2025 02:59 pm IST
FILE PHOTO: Apple announced that Meta’s chief legal officer, Jennifer Newstead, will join the company as its general counsel and senior vice president.
| Photo Credit: Reuters
Apple on Thursday announced that Meta Platforms’ chief legal officer, Jennifer Newstead, will join the company as its general counsel and senior vice president.
Newstead, who previously served as the legal adviser of the U.S. Department of State, will become Apple’s general counsel in March, after assuming the role of senior vice president next month.
The appointment, along with various other executive transitions, follows longtime Apple head of human interface design Alan Dye being appointed as chief design officer at Meta.
Apple’s current general counsel, Kate Adams, who has been in the role since 2017, will retire late next year following a transition of duties, the company said.
The company also announced that Lisa Jackson, who serves as vice president for environment, policy, and social initiatives, will retire in late January.
Apple’s government affairs organization, which falls under its environment, policy, and social initiatives, will merge with the general counsel role and will be led by Newstead after Adams’ retirement.
Published – December 05, 2025 10:52 am IST
FILE PHOTO: Amazon touted new capabilities for its SageMaker AI service that it says will help customers expedite customizing their AI software models.
| Photo Credit: Reuters
Amazon.com on Wednesday touted new capabilities for its SageMaker AI service that it says will help customers expedite customizing their artificial intelligence software models.
For Peter Nikoloff, a different type of sage-making took precedence. He was sure to be first in line Wednesday morning at the most popular booth at Amazon’s annual AWS cloud computing confab in Las Vegas that attracts some 60,000 attendees, paying up to $2,100 each for tickets.
Nikoloff, 37, waited nearly two hours to be first in line to make a custom fragrance for his wife using Amazon’s Nova generative AI software. The voice-powered service spun a fragrance with hints of mint, sandalwood and, yes, sage.
“Oh it was totally worth it,” said Nikoloff, who was attending the weeklong conference, put on by Amazon Web Services, to pick up tips to help with his information technology job with the federal government. “There’s quite a lot of perfume here, for free.”
Sessions dispensing advice on arcane fields such as cloud modernization, data lakes and container migrations are popular among software engineers at the conference, but none has drawn a crowd like the AI perfumery. On the third day of the event, the line for the Fragrance Lab had not abated, drawing upwards of six dozen hopefuls at a time and waits of one hour or more.
The scent lab is a gimmick, but meant to demonstrate the broad applications for AI software, Amazon said. Users are guided through a series of voice prompts on a screen that relies on four different AI models — all Amazon’s — and a scent is generated for human perfumers to prepare. Thirty scents ranging from coffee to tobacco to jasmine can be combined for the final output.
Amazon prompts users to describe their favorite activities, preferred environments and other characteristics before spitting out scents that are meant to match the user’s mood and fragrance profile. A promotional website promises a “uniquely curated” aroma with “rich base notes” or “ethereal top notes.”
Recently mixed scents included Alpine Reverie, Terra Venture, Metropol and Tranquil Pulse, all mixed by (human) perfumers, shipped in from France. Tech firms have boasted of nearly innumerable benefits from the broad rollout of AI, including the speedy creation of new startups, supercharged research reports and the automation of painstaking rote tasks. The conference this year is being used as a platform to demonstrate Amazon’s prowess in AI amid some market concerns that it is lagging rivals.
Since generative AI burst onto the scene about three years ago, it has been deployed to help write novels from scratch, craft movies and commercials, replace engineers and advise budding investors on the market, among thousands of other uses.
AI perfumes, however, should probably be kept in the bottle, said Anthony Walker. He waited about three hours in total on Wednesday for his Amazon Nova-designed cologne but said the output really stinks.
“It was pretty unimpressive: It just didn’t come out as I expected,” said Walker, a 30-year-old engineer working in nuclear power. He had described himself to the AI as nature-loving and creative, but the resulting bamboo- and mint-based aroma named “Sylvan Craft” was, he said, “very feminine.”
“I like more earthy scents or nutty scents,” he said. “I’ll give this to my girlfriend, she’ll be happy with it.”
Published – December 05, 2025 11:05 am IST
FILE PHOTO: Google announced that the Gemini 3 Deep Think mode is being rolled out to Google AI Ultra subscribers in the Gemini app.
| Photo Credit: Reuters
Google announced that the Gemini 3 Deep Think mode is being rolled out to Google AI Ultra subscribers in the Gemini app. The AI model helps users solve complicated math, science and logic problems, taking even a few minutes to respond.
The version is currently the highest-performing AI model on the ARC-AGI-2 reasoning benchmark.
In a blog posted by the company, Google explained this was because it uses advanced parallel reasoning to “explore multiple hypotheses simultaneously.”
The previous variant, Gemini 2.5 Deep Think, which was launched in August, recently achieved a gold-medal standard at the International Mathematical Olympiad and the International Collegiate Programming Contest World Finals.
Google AI Ultra subscribers can access the Deep Think version by choosing ‘Deep Think’ and Gemini 3 Pro in the model dropdown menu.
Published – December 05, 2025 12:27 pm IST
The bill comes as the Trump administration mulls greenlighting sales of Nvidia’s H200 artificial intelligence chips to China, Reuters reported [File]
| Photo Credit: REUTERS
A bipartisan group of U.S. senators, including prominent Republican China hawk Tom Cotton, on Thursday unveiled a bill that would block the Trump administration from loosening rules that restrict Beijing’s access to artificial intelligence chips from Nvidia and AMD for 2.5 years.
The bill, known as the SAFE CHIPS Act, was filed by Republican Senator Pete Ricketts and Democrat Chris Coons. It would require the Commerce Department, which oversees export controls, to deny any license requests for buyers in China, Russia, Iran or North Korea to receive U.S. AI chips more advanced than the ones they currently are allowed to obtain for 30 months. After that, Commerce would have to brief Congress on any proposed rule changes a month before they take effect.
The legislation, which was co-sponsored by Republican Dave McCormick and Democrats Jeanne Shaheen and Andy Kim, represents a rare effort led in part by Trump’s own party to stop him from further relaxing tech export restrictions on China.
“Denying Beijing access to (the best American) AI chips is essential to our national security,” Ricketts said in a statement. The bill comes as the Trump administration mulls greenlighting sales of Nvidia’s H200 artificial intelligence chips to China, Reuters reported. China hawks in Washington fear that Beijing could use the prized chips to supercharge its military with AI-powered weapons and more powerful intelligence and surveillance capabilities.
Faced with new Chinese export curbs on the rare earth metals that global tech companies rely on, Trump’s Commerce Department imposed and then rolled back curbs on Nvidia’s H20 AI chips, a move that was criticized by Republican Representative John Moolenaar, who chairs the House China Select Committee. Advanced Micro Devices, a Nvidia rival, is also eager to sell to China.
As part of negotiations with China to delay its own rare earth controls, Trump pushed back by a year a rule to restrict U.S. tech exports to units of already-blacklisted Chinese companies and has vowed to nix a Biden-era rule restricting AI chip exports globally to countries based in part on concerns around chip smuggling to China.
Greg Allen, a senior adviser at the Center for Strategic and International Studies, a Washington, D.C.-based think tank, called the bill a common-sense measure that should be passed urgently, noting that the United States cannot dissuade China from seeking to swiftly end its reliance on U.S. technology.
“The only choice for America is whether or not we should sell China the technology to make their decoupling strategy fast and convenient,” he said.
Published – December 05, 2025 11:28 am IST