The blog OMG Ubuntu notes that Microsoft Copilot chatbot support has been added in the latest Firefox Nightly builds. "Firefox's sidebar already offers access to popular chatbots, including OpenAI's ChatGPT, Anthropic's Claude, Le Chat's Mistral and Google's Gemini. It previously offered HuggingChat too." As the testing bed for features Mozilla wants to add to stable builds (though not all make it — eh, rounded bottom window corners?), this is something you can expect to find in a future stable update... Copilot in Firefox offers the same features as other chatbots: text prompts, upload files or images, generate images, support for entering voice prompts (for those who fancy their voice patterns being analysed and trained on). And like those other chatbots, there are usage limits, privacy policies, and (for some) account creation needed. In testing, Copilot would only generate half a summary for a webpage, telling me it was too long to produce without me signing in/up for an account.

On a related note, Mozilla has updated stable builds to let third-party chatbots summarise web pages when browsing (in-app callout alerts users to the 'new' feature). Users yet to enable chatbots are subtly nudged to do so each time they right-click on web page. [Between "Take Screenshot" and "View Page Source" there's a menu option for "Ask an AI Chatbot."] Despite making noise about its own (sluggish, but getting faster) on-device AI features that are privacy-orientated, Mozilla is bullish on the need for external chatbots.
The article suggests Firefox wants to keep up with Edge and Chrome (which can "infuse first-party AI features directly.") But it adds that Firefox's nightly build is also testing some non-AI features, like new task and timer widgets on Firefox's New Tab page. And "In Firefox Labs, there are is an option to enable JPEG XL support, a super-optimised version of JPEG that is gaining traction (despite Google's intransigence).

Other Firefox news:

• There's good news "for users still clinging to Windows 7," writes the Register. Support for Firefox Extended Support Release 115 "is being extended until March 2026."

• Google "can keep paying companies like Mozilla to make Google the default search engine, as long as these deals aren't exclusive anymore," reports the blog It's FOSS News. (The judge wrote that "Cutting off payments from Google almost certainly will impose substantial — in some cases, crippling — downstream harms to distribution partners..." according to CNBC — especially since the non-profit Mozilla Foundation gets most of its annual revenue from its Google's search deal.)

• Don't forget you can now search your tabs, bookmarks and browsing history right from the address bar with keywords like @bookmarks, @tabs, and @history. (And @actions pulls up a list of actions like "Open private window" or "Restart Firefox").

At CES in January, Nvidia's CEO Jensen Huang "stood flanked by 14 humanoid robots from different companies," remembers the Washington Post. But how close are we to real-world robot deployments?

Agility Robotics "says its factory is designed to eventually manufacture 10,000 robots a year," the Post adds (with "some" of its robots "already at work in e-commerce warehouses and auto parts factories.") Amazon even invested $150 million in the 10-year-old company (spun out from Oregon State University's robotics lab) in 2022, according to the article, "and has tested the company's robots in its warehouses." The e-commerce revolution has spawned sprawling warehouses across the country where products must be organized and customer orders assembled and shipped, but some human workers have said the repetitive work is low paid and leaves them prone to injury. Agility rents out its robots to warehouse owners it says have struggled to keep their human jobs filled, including logistics company GXO, which uses them at a warehouse for Spanx shapewear in Flowery Branch, Georgia, northeast of Atlanta. The robots pick up baskets of clothing from wheeled robots and walk them over to conveyor belts that take them to other parts of the facility.

Agility Chief Business Officer Daniel Diez said facilities like this represent a first step for humanoid robots into gainful employment. "This work gets paid, and we have eyes on large-scale deployments just doing this, and that's what we're focused on," he said. German auto parts company Schaeffler uses Agility robots to load and unload equipment at a factory in Cheraw, South Carolina. Auto part plants have become a favored proving ground for humanoid robots, with Boston Dynamics, the company famous for its videos of back-flipping robots, doing tests with its majority owner, Hyundai.
But meanwhile, RoboForce makes a robot that has two arms on a base with four wheels, the article notes, "providing stability and making it possible to lift more weight than a bipedal robot." Humanoid designs make sense "if it is so important to justify the trade-off and sacrifice of other things," RoboForce CEO Leo Ma tells the Post. "Other than that, there is a great invention called wheels."

Still, the article argues there's "a new drive to make humanoid robots practical," fueled by "the surge of investment in AI" combined with advancements in robotics that "make humanoid designs more capable and affordable." Years of steady progress have made legged robots better at balancing and stepping through tricky terrain. Improved batteries allow them to operate for longer without trailing industrial power cords. AI developers are adapting the innovations behind services like ChatGPT to help humanoids act more independently... The progress has triggered a frenzy of investment in humanoid robots and made them into a mascot for the idea that AI will soon reorder the world on the scale tech leaders have promised... Venture capitalists have invested over $5 billion in humanoid robotics start-ups since the beginning of 2024, according to financial data firm Pitchbook, and the largest tech corporations are also placing bets... Meta is working on integrating its own AI technology with humanoid robots, and Google researchers are collaborating with Austin-based humanoid robot start-up Apptronik... A host of humanoid robot companies has spawned in China, the world leader in complex manufacturing, where the government is subsidizing the industry. Six of the 14 robots that shared the stage with Nvidia's Huang were made by Chinese companies; five were American.
"China's Unitree sells a 77-pound humanoid that stands 4-foot-3 for $16,000..."

"A U.S.-based nuclear fusion developer wants to deploy a reactor in Japan in the late 2030s or early 2040s," reports Bloomberg, "in line with the Asian country's broader plans to adopt the potent, low-carbon energy source." Commonwealth Fusion Systems, which last week announced it raised $863 million from investors including Nvidia, has been in dialogue with Japanese government officials on the use of its technology, CEO Bob Mumgaard said in an interview in Tokyo on Wednesday... Several countries are eyeing the technology for its climate and energy security benefits but only some, like China, the U.S., Russia and South Korea have managed to crack the basics. Japan revised its national strategy in June to support fusion deployment and build a demonstration plant in the 2030s.
The article notes that Commonwealth "does not currently have any reactors in operation" — but that Mitsubishi this week invested in the company, in collaboration with a consortium of 12 Japanese companies. From Mitsubishi's announcement: The Japanese Consortium will acquire technical and commercial expertise in policy, regulatory, and the development, construction, operation, and maintenance of ARC [power plant] from CFS's commercialization projects in the United States. In addition, each consortium company will bring together its know-how and expertise and aspire to expedite the commercialization and industrialization of fusion energy power generation in Japan.

"A federal jury has ordered Google to pay $425.7 million for improperly snooping on people's smartphones during a nearly decade-long period of intrusions," reports the Associated Press: The lawyers who filed the case had argued Google had used the data they collected off smartphones without users' permission to help sell ads tailored to users' individual interests — a strategy that resulted in the company reaping billions in additional revenue. The lawyers framed those ad sales as illegal profiteering that merited damages of more than $30 billion. Even though the jury came up with a far lower calculation for the damages, one of the lawyers who brought the case against Google hailed the outcome as a victory for privacy protection. "We hope this result sends a message to the tech industry that Americans will not sit idly by as their information is collected and monetized against their will," said attorney John Yanchunis of law firm Morgan & Morgan.
David Boies, the man who led the U.S. government's 2001 antitrust prosecution of Microsoft, was the plaintiffs' attorney. More details from Bloomberg Law: The lawsuit alleged that since 2016 Google told its users that when they turned off a privacy setting known as Web & App Activity, the company would cease collecting their data from third-party apps that use Google's back end data analytics services. Google continued that collection despite its promise to users that they had control, the plaintiffs alleged. Judge Richard Seeborg certified a class of 98 million Google users who has switched the Web & App Activity setting off...

Boies told the jury during closing statements that the case was about Google breaking its promise to users that they had control over their data. He pointed to Congressional testimony from Google CEO Sundar Pichai in 2018 who said users could clearly see what information the company had, all while internal communications and surveys said users were being misled about their privacy... During closing statements, Google attorney Benedict Hur of Cooley LLP said that as soon as a user click the tracking switch off, they were presented with an "Are You Sure?" screen that stated that users can "learn about the data Google continues to collect and why" by clicking an additional link.
A spokesperson for Google said they would appeal the verdict.

Microsoft is giving away Microsoft 365 Personal subscriptions to all US college students. From a report: This subscription gives students free access to Microsoft's Office apps and the Copilot AI assistant integration for a year, after which the students are eligible for a 50 percent discount to continue the subscription.

While most students have access to education versions of Microsoft 365 or Google Workspace, Microsoft's offer is for student's own personal Microsoft accounts, and is available to claim until October 31st. Microsoft 365 Personal is usually $99.99 a year, or $9.99 a month, and includes 1TB of OneDrive cloud storage.

Nobel laureate Geoffrey Hinton has warned that AI will concentrate wealth among a small elite while impoverishing most workers. The computer scientist, who pioneered neural network research in the 1980s, told Financial Times that rich people will use AI to replace workers, creating massive unemployment and profit increases.

Hinton, who left Google in 2023 after selling his AI startup for $44 million a decade earlier, dismissed universal basic income as insufficient to address human dignity concerns from job losses. The 77-year-old physicist predicts superintelligent AI will arrive within five to twenty years. He blamed capitalism rather than AI technology itself for the coming economic disruption, stating the system ensures AI will primarily benefit the wealthy rather than solve grand problems like hunger or poverty.

Alphabet's Google was hit with a $3.45 billion EU antitrust fine on Friday for anti-competitive practices in its lucrative adtech business, marking its fourth penalty in its decade long fight with EU competition regulators. From a report: The move by the European Commission was triggered by a complaint from the European Publishers Council and comes amid a threat by U.S. President Donald Trump to retaliate against the European Union for any push against Big Tech.

The EU competition enforcer had originally planned to hand out the fine on Monday but opposition from EU trade chief Maros Sefcovic on concerns about the impact on U.S. tariffs on European cars derailed EU antitrust chief Teresa Ribera's plan. The Commission said Google favored its own online display technology services to the detriment of rivals and online publishers and that it abused its market power since 2014 until today.

An anonymous reader shares a report: Google's CEO Sundar Pichai stood smiling in a leafy-green California garden in September 2020 and declared that the tech behemoth was entering the "most ambitious decade yet" in its climate action. "Today, I'm proud to announce that we intend to be the first major company to operate carbon free -- 24 hours a day, seven days a week, 365 days a year," he said, in a video announcement at the time.

Pichai added that he knew the "road ahead would not be easy," but Google "aimed to prove that a carbon-free future is both possible and achievable fast enough to prevent the most dangerous impacts of climate change." Five years on, just how hard Google's "energy journey" would become is clear. In June, Google's Sustainability website proudly boasted a headline pledge to achieve net-zero emissions by 2030. By July, that had all changed. An investigation by Canada's National Observer has found that Google's net-zero pledge has quietly been scrubbed, demoted from having its own section on the site to an entry in the appendices of the company's sustainability report.

Atlassian said it has agreed to acquire The Browser Co., a startup that offers a web browser with AI features, for $610 million in cash. CNBC: The companies aim to close the deal in Atlassian's fiscal second quarter, which ends in December. Established in 2019, The Browser Co. has gone up against some of the world's largest companies, including Google, with Chrome, and Apple, which includes Safari on its computers running MacOS. The startup debuted Arc, a customizable browser with a built-in whiteboard and the ability to share groups of tabs, in 2022.

The Dia browser, a simpler option that allows people to chat with an AI assistant about multiple browser tabs at once, became available in beta in June. Atlassian co-founder and CEO Mike Cannon-Brookes said he sees shortcomings in the most popular browsers for those who do much of their work on computers. Further reading: Atlassian Buying The Browser Company Feels Like a Waste of Money.

Google's September Pixel drop brings the new Material 3 Expressive UI, AI-powered Gboard writing tools, and Bluetooth Auracast upgrades to older Pixel devices, including the Pixel 6 and Pixel Tablet. "Among other tweaks, Google made it possible to add 'Live Effects,' including a few that cover the weather, to your phone's lock screen wallpaper," notes Engadget. "Material 3 Expressive also gives you more control over how the contact cards your phone displays when your friends and family call you look. Even if you're not one to endlessly tweak Android's appearance, as part of the redesign Google has once again reworked the Quick Settings pane in hopes of making it easier to use."

On the audio front, Pixel Buds Pro 2 gain intuitive nod-and-shake gesture controls, Adaptive Audio for balanced awareness, and Loud Noise Protection to guard against sudden sound spikes. Voice clarity has also been improved with Gemini Live in noisy environments.

A full breakdown of what's new can be found here.

An anonymous reader quotes a report from ZDNet: Over the Labor Day weekend, Cloudflare says it successfully stopped a record-breaking distributed denial-of-service (DDoS) attack that peaked at 11.5 terabits per second (Tbps). This came only a few months after Cloudflare blocked a then all-time high DDoS attack of 7.3 Tbps. This latest attack was almost 60% larger.

According to Cloudflare, the assault was the result of a hyper-volumetric User Datagram Protocol (UDP) flood attack that lasted about 35 seconds. During that just more than half-minute attack, it delivered over 5.1 billion packets per second. This attack, Cloudflare reported, came from a combination of several IoT and cloud providers. Although compromised accounts on Google Cloud were a major source, the bulk of the attack originated from other sources.

The specific target of this attack has not been publicly disclosed, but we can be sure the intent was to overwhelm the victim's network and render online services inoperative. Cloudflare says its globally distributed, fully autonomous DDoS mitigation network detected and neutralized the threat in real time, without notable impact on customer services or requiring manual intervention. This operation highlights both the rising sophistication of attack methods and the resilience of modern internet infrastructure defenses, especially Cloudflare's use of real-time packet analysis, fingerprinting, and rapid threat intelligence sharing across its network.

Critics are denouncing Tuesday's antitrust remedies ruling against Google, calling them inadequate to restore search market competition. DuckDuckGo said the court's decision allows Google to continue using its monopoly to hold back competitors in AI search.

The Open Markets Institute called it "pure judicial cowardice" that leaves Google's power "almost fully intact." Senator Amy Klobuchar said the limited remedies demonstrate why Congress needs to pass legislation stopping dominant platforms from preferencing their own products. The News/Media Alliance criticized Judge Amit Mehta for failing to address Google forcing publishers to provide content for AI offerings to remain in search results.

Chrome has extended its dominance in the browser wars, surpassing 70% market share on desktops while Edge, Safari, Firefox, and Opera trail far behind. Neowin reports: According to [Statcounter], in August 2025, Chrome kept on increasing its overwhelming market share, which is now above the 70% mark (70.25%, to be precise) in the desktop browser market. The gap between Chrome and its closest competitor, Microsoft Edge, is immense, with Edge holding just 11.8% (+0.01 points over the previous month). Apple's Safari is third with 6.34% (+1.04 points); Firefox has 4.94% (-0.36 points); and Opera is fifth with a modest 2.06% market share (-0.13 points).

Things look similar on the mobile side of the market, with Google Chrome having 69.15% (+1.92 points) and Safari being second with 20.32% (-2.2 points). Samsung Internet is third with 3.33% (-0.17 points). As for Microsoft Edge, its mobile share is only 0.59% (+0.06 points). The findings can be found here.

A federal judge spared Google from the harshest penalties in its antitrust case. The search giant can keep Chrome and avoid breaking up Android, but it has been barred from exclusive contracts and ordered to limit data sharing with rivals. CNBC reports: U.S. District Judge Amit Mehta ruled against the most severe consequences that were proposed by the U.S. Department of Justice, including selling off its Chrome browser, which provides data that helps its advertising business deliver targeted ads. "Google will not be required to divest Chrome; nor will the court include a contingent divestiture of the Android operating system in the final judgment," the decision stated. "Plaintiffs overreached in seeking forced divesture of these key assets, which Google did not use to effect any illegal restraints."

The company can make payments to preload products, but it cannot have exclusive contracts, the decision stated. The DOJ asked Google to stop the practice of "compelled syndication," which refers to the practice of making certain deals with companies to ensure its search engine remains the default choice in browsers and smartphones. [...] The judge ordered the parties to meet by September 10th for the final judgement.

"Google will not be barred from making payments or offering other consideration to distribution partners for preloading or placement of Google Search, Chrome, or its GenAI products. Cutting off payments from Google almost certainly will impose substantial -- in some cases, crippling -- downstream harms to distribution partners, related markets, and consumers, which counsels against a broad payment ban." [...] Google said it will appeal the ruling, which would delay any potential penalties. Mehta ruled Tuesday that Google will have to make available certain search index data and user interaction data though "not ads data." The court narrowed the datasets Google will be required to share and said they must occur on "ordinary commercial terms that are consistent with Google's current syndication services."

Developer Hugo Tunius, writing in a blog post: Sideloading has been a hot topic for the last decade. Most recently, Google has announced further restrictions on the practice in Android. Many hundreds of comment threads have discussed these changes over the years. One point in particular is always made: "I should be able to run whatever code I want on hardware I own." I agree entirely with this point, but within the context of this discussion it's moot.

When Google restricts your ability to install certain applications they aren't constraining what you can do with the hardware you own, they are constraining what you can do using the software they provide with said hardware. It's through this control of the operating system that Google is exerting control, not at the hardware layer. You often don't have full access to the hardware either and building new operating systems to run on mobile hardware is impossible, or at least much harder than it should be. This is a separate, and I think more fruitful, point to make. Apple is a better case study than Google here. Apple's success with iOS partially derives from the tight integration of hardware and software. An iPhone without iOS is a very different product to what we understand an iPhone to be. Forcing Apple to change core tenets of iOS by legislative means would undermine what made the iPhone successful.

Google denied claims Monday that it had issued a security warning to Gmail users about a major vulnerability. The company stated that recent reports claiming a broad Gmail security alert were "entirely false." Google said its email service blocks more than 99.9% of phishing and malware attempts from reaching users' inboxes.

Waymo robotaxis are repeatedly selecting identical parking spots in front of specific Los Angeles and Arizona homes between rides, puzzling residents who document the same vehicles returning to precise locations daily. The company states its vehicles choose parking based on local regulations, existing vehicle distribution, and proximity to high-demand areas but cannot explain the algorithmic specificity.

Carnegie Mellon autonomous vehicle expert Phil Koopman attributes the behavior to machine learning systems optimizing for specific spots without variation. Waymo said it had received neighbor complaints and has designated certain locations as no-parking zones for its fleet. The vehicles comply with three-hour parking limits, according to Los Angeles Department of Transportation regulations, governing commercial passenger vehicles under 22 feet.

"AI web crawlers are strip-mining the web in their perpetual hunt for ever more content to feed into their Large Language Model mills," argues Steven J. Vaughan-Nichols at the Register.

And "when AI searchbots, with Meta (52% of AI searchbot traffic), Google (23%), and OpenAI (20%) leading the way, clobber websites with as much as 30 Terabits in a single surge, they're damaging even the largest companies' site performance..." How much traffic do they account for? According to Cloudflare, a major content delivery network (CDN) force, 30% of global web traffic now comes from bots. Leading the way and growing fast? AI bots... Anyone who runs a website, though, knows there's a huge, honking difference between the old-style crawlers and today's AI crawlers. The new ones are site killers. Fastly warns that they're causing "performance degradation, service disruption, and increased operational costs." Why? Because they're hammering websites with traffic spikes that can reach up to ten or even twenty times normal levels within minutes.

Moreover, AI crawlers are much more aggressive than standard crawlers. As the InMotionhosting web hosting company notes, they also tend to disregard crawl delays or bandwidth-saving guidelines and extract full page text, and sometimes attempt to follow dynamic links or scripts. The result? If you're using a shared server for your website, as many small businesses do, even if your site isn't being shaken down for content, other sites on the same hardware with the same Internet pipe may be getting hit. This means your site's performance drops through the floor even if an AI crawler isn't raiding your website...

AI crawlers don't direct users back to the original sources. They kick our sites around, return nothing, and we're left trying to decide how we're to make a living in the AI-driven web world. Yes, of course, we can try to fend them off with logins, paywalls, CAPTCHA challenges, and sophisticated anti-bot technologies. You know one thing AI is good at? It's getting around those walls. As for robots.txt files, the old-school way of blocking crawlers? Many — most? — AI crawlers simply ignore them... There are efforts afoot to supplement robots.txt with llms.txt files. This is a proposed standard to provide LLM-friendly content that LLMs can access without compromising the site's performance. Not everyone is thrilled with this approach, though, and it may yet come to nothing.

In the meantime, to combat excessive crawling, some infrastructure providers, such as Cloudflare, now offer default bot-blocking services to block AI crawlers and provide mechanisms to deter AI companies from accessing their data.

After Will Smith uploaded a video of an adoring crowd, blogger Andy Baio "conducted a detailed analysis that suggests Will Smith's team might have used AI to turn photos from his recent concerts into videos," writes BGR. But there's more to the story: Google recently ran an experiment for YouTube Shorts in which it used AI (machine learning) to improve the quality of Shorts without asking the creator for permission. People complained the videos looked like they were AI generated. It seems that Will Smith's YouTube Shorts clip that attracted criticism from fans this week might have been a victim of this experiment... The signs are real. The man who claimed Will Smith's song helped him cure cancer was there. The woman in front of him was holding the sign with him. The "Lov U" sign appeared in photos the singer posted on his social media channels before the clip was shared.
"Will Smith has not denied the use of AI in these promotional clips," the article adds.

But the Hollywood Reporter also calls it "just the beginning of AI chaos," noting that "influencers and spinmeisters have been using AI upscaling for years, if quietly, the way you might round up your current salary in a job interview." It's only going to grow more popular as the tools get better. (And they will — you just need some tweaks to the model and increases in compute to erase these hallucinations.) In fact, when the chapter on the early AI Age is written, the line about this moment is less likely to be, "Remember when Will Smith did something cringily AI?" and more, "Remember when AI was still seen as so cringe that we made fun of Will Smith for it?" Experts differ on the timeline, but everyone agrees it's just years if not months before we'll stop being able to spot an AI video. [Will Smith's video] had the particular misfortune of coming out at this interregnum moment: good enough for someone to use but not so good we can't spot it.

That moment will be over soon enough, and, I suspect, so will our pearl-clutching. The main effect of this new age of the synthetic is that video will stop being a meaningful measure of truth. We have long stopped believing everything we read, and AI image-generators have killed what photoshop wounded. But video until now has been the last bastion of objectivity — incontrovertible evidence that an event took place the way it seemed to....

But there is an upside. (Really.) Without a format that can telegraph objectivity, we'll need to (if we care to) turn to other ways to assure ourselves of the facts: the source of the video. That could mean the human-led content creator will matter more. After years of seeing news brands take a beating in the trust department, they'll soon become the only hope we have of knowing whether something happened. We no longer will be able to trust the medium. But we may newly believe the media.

When it comes to nuclear fusion energy, "How do we advance fusion as fast as possible?" asks the CEO of Commonwealth Fusion Systems. They've just raised $863 million from Nvidia, Google, the BIll Gates-founded Breakthrough Energy Ventures and nearly two dozen more investors, which "may prove helpful as the company develops its supply chain and searches for partners to build its power plants and buy electricity," reports TechCrunch.

Commonwealth's CEO/co-founder Bob Mumgaard says "This round of capital isn't just about fusion just generally as a concept... It's about how do we go to make fusion into a commercial industrial endeavor." The Massachusetts-based company has raised nearly $3 billion to date, the most of any fusion startup. Commonwealth Fusion Systems (CFS) previously raised a $1.8 billion round in 2021...

CFS is currently building a prototype reactor called Sparc in a Boston suburb. The company expects to turn that device on later next year and achieve scientific breakeven in 2027, a milestone in which the fusion reaction produces more energy than was required to ignite it. Though Sparc isn't designed to sell power to the grid, it's still vital to CFS's success. "There are parts of the modeling and the physics that we don't yet understand," Saskia Mordijck, an associate professor of physics at the College of William and Mary, told TechCrunch. "It's always an open question when you turn on a completely new device that it might go into plasma regimes we've never been into, that maybe we uncover things that we just did not expect." Assuming Sparc doesn't reveal any major problems, CFS expects to begin construction on Arc, its commercial-scale power plant, in Virginia starting in 2027 or 2028...

"We know that this kind of idea should work," Mordijck said. "The question is naturally, how will it perform?" Investors appear to like what they've seen so far. The list of participants in the Series B2 round is lengthy. No single investor led the round, and a number of existing investors increased their stakes, said Ally Yost, CFS's senior vice president of corporate development... The new round will help CFS make progress on Sparc, but it will not be enough to build Arc, which will likely cost several billion dollars, Mumgaard said.
"As advances in computing and AI have quickened the pace of research and development, the sector has become a hotbed of startup and investor activity," the article points out.

And CEO Mumgaard told TechCrunch that their Sparc prototype will prove the soundness of the science — but it's also important to learn "the capabilities that you need to be able to deliver it. It's also to have the receipts, know what these things cost!"