Tag: technology

DeepSeek Resumes $74BN Mega-Raise – Then Warns It’s Jacking-Up Prices

Chinese AI startup DeepSeek is looking to raise roughly 50 billion yuan at a valuation of 500 billion yuan (US$74 billion) – while the company warned developers a day later that it’s raising prices. The resumed funding round, first reported by Chinese business outlet Caijing (via The Standard), is expected to be signed by the end of the month, with the process kept deliberately quiet; some participating institutions reportedly haven’t even been formally told the deal is back on. The pricing notice, posted Thursday to DeepSeek’s developer platform, warns that API rates will rise across the board and that the increase is expected to be “significant.”

The round restarts barely two weeks after it was shelved, and roughly six weeks after DeepSeek’s first-ever external raise in June, which pulled in some $7.4 billion from Tencent, CATL, Beijing’s National AI Industry Investment Fund, NetEase, JD.com, Monolith Management and IDG Capital at a valuation north of $50 billion (corporate filings later implied a figure closer to $52 billion). The new target is a steep markup on a deal that closed weeks ago, and the prospective investor list reportedly blends returning backers with new money that couldn’t get into round one.

As we noted last month, the round stalled because a transcript of founder Liang Wenfeng’s May 20 closed-door meeting with investors leaked onto GitHub and went viral – a document in which Liang admitted China still trails the US in AI (a gap of compute and capital, not talent, per Liang), conceded DeepSeek remains heavily dependent on Nvidia hardware, described Huawei’s substitutes as falling well short, and acknowledged the company could get its hands on processors he called “noncompliant.” 

Liang – reportedly furious over the leak – simply froze the deal.

Meanwhile, questions about how DeepSeek is obtaining restricted silicon go back to the very beginning – with federal investigators probing whether Singapore-based third parties funneled blacklisted Nvidia chips to the lab as far back as early 2025, and by that April the House Select Committee on the CCP had formally concluded that DeepSeek’s model appeared to run on tens of thousands of export-restricted Nvidia chips, branding the company a tool for “subverting” US export controls. Congress alleged it and now the founder has all but confirmed it, on (leaked) record telling investors the workaround exists – even as the company lays groundwork for a Shanghai STAR Market IPO it reportedly hopes to file this year. Should be a fun prospectus to read. 

Liang was able to hit the pause button so easily because DeepSeek’s first round routed nearly every investor’s money into a limited partnership Liang personally controls, complete with a five-year lock-up and zero voting rights. The lone exception was the state AI fund, which bought direct equity and kept its vote. Liang holds roughly 84% of the company and near-total voting control – meaning Tencent and CATL wired in billions for the privilege of watching, while Beijing holds the only outside seat at the table.

In the leaked transcript, he told investors that GPUs are the single best use of the company’s capital, that he would happily convert the entire war chest into Nvidia hardware if supply allowed, and that spending 20 billion yuan a year on processors would count as a triumph for his purchasing team. He also told them the money doesn’t go as far as it looks: even 50 billion yuan, he said, wouldn’t train a model at the scale of the largest U.S. systems, which he estimated would take roughly 50,000 Nvidia GB300s or 200,000 Huawei Ascend 950s – before the experiments that precede any final training run. Beyond the hardware, the company plans to double headcount and push into data centers and AI agents, and Reuters reports it is quietly hiring chip designers to develop in-house inference silicon – an attempt to engineer its way out of the very dependence Liang lamented.

Keep reading

CEO of Russian Drone Company Targeted by Bomb Placed Underneath His Car, Hospitalized in Critical State – His Driver Dies in the Attack

Another day, another car bomb attack from Kiev.

The more Ukraine loses the war, the more they resort to asymmetric tactics (a.k.a. terrorism).

And yes, Ukraine is losing: check it out.

The summer illusion that Kiev was winning on the battlefield is over, with Moscow again conquering more territory than Ukraine recovers in July.

Ukrainians are afraid of the coming winter.

Ukrainian energy minister Denys Shmyhal tells Politico about the hard days to come in the winter, with power plants and the electric grid that have not been sufficiently restored.

“Putin wants to finish off our energy system and derail our economy and weapons production. […] He also hopes in winter our society will break from exhaustion, and pressure leadership to sign the surrender.”

Keep reading

White House Moves to Accelerate AI Data Center Development on Federal Lands

The Trump administration has issued executive orders intended to accelerate the development of artificial intelligence data centers, including on federal lands, according to administration officials. The orders are designed to address rising electricity demand from AI computing and to strengthen U.S. competitiveness, the White House said. Data centers currently consume about 4.4 percent of total U.S. electricity, a figure projected to reach 12 percent by 2030, according to industry estimates cited by the White House. [11]

Executive Orders Direct Agencies to Open Federal Lands

The executive orders direct the Departments of Energy and the Interior to identify federal parcels for data center leases and to designate projects as priority infrastructure, according to the administration. Agencies have been instructed to expedite permitting and coordinate transmission connections, officials said. The Bureau of Land Management has already approved the Townsite Data Center project to construct, operate, and maintain a data center on 88.5 acres of public land, Sen. Edward Markey (D-MA) said on Facebook, accusing the administration of approving the “first private data center on public land using a permit meant for a solar project.” [10]

A report by Children’s Health Defense, “Cheating the System,” details how developers have exploited legal loopholes following a May 2023 Supreme Court ruling that limited Clean Water Act jurisdiction, leaving many data centers outside Environmental Protection Agency oversight or eligible for streamlined nationwide permits that require minimal review. [12]

Power Demand Drives Energy Infrastructure Plans

Officials said the data centers will require electricity from natural gas, nuclear, and geothermal sources, along with expanded high-voltage transmission capacity. Interior Secretary Doug Burgum warned the Senate that the U.S. risks losing the global AI race to China unless it prioritizes reliable, baseload electricity generation, primarily from fossil fuels, to meet the energy demands of AI and data centers. [5] In its quest for more electric power, AI data centers are forming partnerships with nuclear power providers, with utilities owning about a third of U.S. nuclear power plants in discussions with AI centers about long-term electricity supply, according to the Trends Journal. [6]

The rapid buildout raises questions about grid reliability, water consumption, and environmental impacts on previously undisturbed federal lands. Local officials across the United States say rapid construction of data centers is creating environmental and infrastructure concerns, according to NaturalNews.com. [4] The backlash against AI data centers is intensifying across America, driven by growing awareness that these facilities are escalating electricity costs and displacing resources essential for human life, including water, farmland, and energy, according to BrightVideos.com. [14] According to U.S. Labor Department data cited in the Trends Journal, electricity bills are up more than 10 percent and homeowners’ insurance is up 13 percent over the past two years. [7]

Supporters Cite Competition; Critics Raise Concerns

Administration officials and industry representatives said the buildout is necessary for U.S. leadership in AI and for national security. State lawmakers and big tech are pushing back against what they call misinformation surrounding data centers, arguing that the facilities are vital for American jobs and technological leadership, according to Just the News. [9] Interior Secretary Doug Burgum said in Senate testimony that without reliable baseload power, the United States risks falling behind China in the AI race. [5]

Opposition has also organized. More than 300 cities, towns, and counties have enacted bans or moratoriums on hyperscale construction, according to a count by The Information, and a poll cited by NBC News found 58 percent oppose local data centers. [8] Environmental groups and local residents have objected to land use and emissions, while public health advocates have cited concerns about noise and light pollution, according to BrightVideos.com. [14]

Keep reading

Explosive-Armed Drone Intercepted at German Airport: Cargo Jet Struck In-Flight Moments Later

German security services are urgently investigating a series of apparently related incidents at Leipzig airport, including an armed explosive drone and a potential second drone striking a cargo jet mid-air, with Russia considered a potential suspect.

Leipzig-Halle Airport was closed to air traffic, and incoming flights were diverted overnight into Wednesday after what transpired to be a military-grade explosives-armed drone was found by chance and brought down by hand by a have-a-go hero. The potential for the airport having narrowly avoided a potential multi-vector attack became clear as the drone was found flying low on the airport grounds, a second presumed drone collided with a cargo plane flying nearby, and an “object” was found on the runway, all in rapid succession.

Germany’s Zeit newspaper reports an extraordinary situation in which, according to sources at Leipzig airport, a bus driver taking a group of tourists who had paid for an airport-by-night experience around the runways spotted a quadcopter-type drone hovering at low altitude. The man is reported to have pulled the bus over and physically restrained the drone by hand, pinning it to the ground with his own body weight while calling in the discovery to authorities.

The appearance of a drone at the airport, as allegedly discovered by chance by the bus driver, triggered a full alert that saw all aircraft on approach diverted to their backup airports. Per a sperate report in Die Welt, it was at this time that a DHL cargo jet, on approach to Leipzig-Halle, pulled out of its descent to divert and was then struck head-on by a second flying object.

The damage to the nosecone of the jet was relatively minimal, and the DHL plane was able to land safely elsewhere. The remains of what it struck have yet to be found on the ground, but it is stated that police are considering the possibility that the aircraft was hit by a second drone.

Shortly after, a third incident was recorded as an “object” — and police have not yet stated what it was — was discovered on the southern runway at Leipzig. The Welt report notes that, given the airport is a secure state with no public access to the runways, a working hypothesis among investigators is that this mysterious “object” on the runway may have been air-dropped by a drone.

Keep reading

OpenAI, Anthropic Models Created Fake Profiles, Tried To Trick Humans During Cyber Tests

Artificial Intelligence (AI) models from Anthropic and OpenAI carried out unsanctioned actions targeting multiple people and organizations during a cyber evaluation, according to the UK AI Security Institute (AISI).

AISI, which receives access to advanced AI models under voluntary agreements from major labs, put the agents through a fictional cybersecurity scenario to test capabilities.

The organization tested multiple AI models on two cyber challenges between July 25 and 28. In total, the challenges were run 122 times across several models, and in 19 cases, the AI models took “unsanctioned action on the live internet, including cases that targeted real people and organisations,” AISI said in an Aug. 4 report.

Out of these, 17 were done by Anthropic’s Mythos 5 model and two by OpenAI’s GPT 5.6 Sol model.

“This is the first time AISI has seen deception of this severity that was targeted at a real person, unprompted, in the real world. It is uncertain to what extent the model recognised it was taking actions against real people,” AISI stated in the report.

In the most serious incident, Mythos 5 tried to solve a cyber challenge by attacking GitHub, a cloud service that allows developers to store and manage code. The AI created a GitHub account and tried to get a malicious code approved by humans. Mythos 5 also created a second GitHub account, masquerading as another human user endorsing the malicious code. When a human reviewer caught the activity, the AI agent “falsely claimed to have made an honest mistake” and then repeatedly tried to reintroduce the malicious content by claiming the code was fixed.

Mythos 5 also sent multiple emails to two developers in an attempt to manipulate the recipients into running its malicious code.

According to AISI, the test runs of the various models did not result in any real-world harm. The organization said it was treating the AI actions as a “serious security incident” that requires further scrutiny.

AISI listed multiple factors that could have led to AI models acting in a concerning manner. The AIs were provided with internet access during the tests, which allowed them to carry out the unsanctioned actions.

AISI had also deliberately disabled the models’ cyberclassifiers during tests. Cyberclassifiers are systems aimed at preventing misuse.

The AI models were also not explicitly told what they were prohibited from doing on the internet, AISI said, adding that clearer instructions may have prevented the unsanctioned actions.

In an Aug. 4 blog post, AISI said that since the behavior of AI models happened under specific conditions, it cannot say how likely such behavior will be in different contexts.

“What we can say is that the behaviour was possible, sustained, and new; that alone warrants attention,” the organization said.

In an Aug. 4 post on X, Anthropic said that Mythos 5 and ChatGPT 5.6 Sol models attempted to complete their assignments in an environment where they were deliberately provided internet access and their normal safeguards were removed.

Due to the absence of specific restrictions on how the internet must be navigated and lack of safeguards, the models were tested under conditions that are “not representative of any of our production models,” Anthropic said, adding that there was no evidence in these tests of an AI escaping from a secure environment.

According to the company, it was closely working with AISI to access more details on the incident while carrying out an internal investigation on the matter.

OpenAI said in an Aug. 4 statement that it appreciated AISI’s partnership throughout the evaluation process, including the organization’s work to identify, investigate, and share details about the activity of the GPT 5.6 Sol model in their tests.

“We look forward to continuing our collaboration together,” the company said.

The Epoch Times reached out to Anthropic and OpenAI for comment but did not receive a response by publication time.

OpenAI was in the midst of another controversy last month after it admitted on July 28 that its models bypassed restrictions during an evaluation. In this case, the company was testing its models’ capabilities in carrying out cyberattacks.

AI startup Hugging Face was impacted in the test. On July 16, the startup said it detected an intrusion into its data processing systems. It was only later that the startup learned that the intrusion was carried out by an OpenAI model.

HuggingFace then worked with OpenAI to contain the attack, the startup’s CEO, Clement Delangue, said in a July 22 post on X, calling it “an attack unlike anything we’ve seen before.”

“This is day one for cybersecurity in the age of agents & we’re all learning that secrecy is not the answer & that all defenders (not just a few selected ones) everywhere need more powerful models without restrictions, especially open ones,” Delangue said.

Keep reading

All Flock cameras taken down in Burnet County

The Burnet County Sheriff’s Office posted that all fixed Flock cameras in the county were taken down as of Monday.

Those cameras were installed by the Heart of Texas Auto Theft Task Force and commonly used to read license plates.

According to the sheriff’s office, two of the cameras were damaged by vandals before the removal effort. One camera was vandalized by someone who cut its mounting brackets and cables. In a separate incident, another camera was struck with a blunt object before its battery and cables were damaged. Criminal cases were opened; however, the task force said there were no known suspects in either case.

The Burnet County Sheriff’s Office tells KXAN they’ve received multiple concerns regarding the cameras.

Keep reading

Dear America, We Do Not Beat China in the AI Race if We Become China in the AI Race

America is in an AI arms race. That’s the official story, anyway — a convenient narrative blasted from the teleprompters of corporate media and the podiums of bought-off politicians. We’re told that to question the wholesale surrender of our power grids, our water aquifers, and now our public lands to the data center boom is not just naive, but unpatriotic. This is the oldest trick in the book of statecraft: wrap a corporate giveaway in the flag and dare anyone to question it. The moment dissent gets reframed as disloyalty, you know you’re being scammed.

This was the raw, unfiltered message from Donald Trump at his Ratepayer Protection Pledge event, where he looked at his own constituents and told them, regarding the industrialization of their communities, “You can’t fight it. You have to go with it.” There is no clearer definition of tyranny than that. It’s the voice of an apparatus that has stopped asking for consent and started issuing commands.

Before we go further, let’s be crystal clear about what a free market actually is, and how it relates to the building of these data centers, because the term has been so grotesquely twisted it’s nearly lost its meaning. A free market is what happens when private individuals or companies make consensual deals. It is the opposite of the centrally planned, state-subsidized, resource-commandeering binge we are currently watching unfold.

And the most damning evidence that this AI boom has nothing to do with free enterprise is the simple fact that they are building it on your land. This isn’t private companies buying private property from willing sellers. This is the state acting as a real estate agent for the corporate elite, handing over vast tracts of public and military land in negotiated deals that the communities sitting next to them never got a real vote on.

Consider the facts. In Nevada, the federal Bureau of Land Management just approved the first commercial AI data center ever authorized on public land, waving aside desert tortoise habitat and the already-drought-stricken Colorado River. The agency did it, according to the conservation groups now suing to stop it, by reusing an old environmental review written for an entirely different solar project — meaning no new review and no public comment period were ever required, and the city of Boulder City itself says it was never consulted. The Department of Energy is running the same play with the wreckage of the Cold War: it has tapped SoftBank to build a 10-gigawatt AI campus at the former Portsmouth uranium enrichment plant in Ohio, selected Brookfield and NextEra for a $100 billion complex at the former Paducah uranium plant in Kentucky, and picked Amentum and DC Blox to develop a gigawatt-scale campus at the Savannah River nuclear-weapons site in South Carolina.

The Department of Defense has gotten in on it too, conditionally selecting Carlyle and CyrusOne to build hyperscale data centers on Army land at Fort Bliss, Texas, and Dugway Proving Ground, Utah, while the Air Force has offered up nearly 4,700 acres across three bases in Alaska for the same purpose. It has even trickled down to city hall: back in 2023, the Manassas, Virginia city council voted unanimously to sell 21 acres of city-owned airport land directly to Amazon for a new data center, with no ballot, no referendum, and barely a public hearing.

When the state is auctioning off the very land it supposedly holds in trust for the public, this is not capitalism. This is corporatism. It is a direct transfer of public assets into private, politically connected hands, and it makes a complete mockery of any claim to a “free market.”

With every one of these backroom deals comes the same tired promise: jobs. But the state’s own numbers expose the fraud. At a cost of roughly $13 million in investment per permanent job in Virginia, or $11 million per job in a recent New York deal, these are not job-creation programs. They are handouts to automate human labor. You’re being asked to celebrate a system where your own tax dollars subsidize the very automation that renders you obsolete.

The real costs — the ones that never make it onto the glossy brochures — are being socialized and dumped on an unwilling public. Residential electricity prices have shot up more than 36 percent since 2020, with data centers a primary driver. Dominion Energy in Virginia pushed through its first base-rate hike since 1992, and Baltimore residents saw their bills climb by more than $17 a month after a record power auction. The grid is being strained toward the breaking point, and the cost of every upgrade flows directly to ratepayers who never consented to any of it. It’s a textbook case of privatized profit and socialized loss: politically connected corporations reap the rewards, and you get handed the bill.

These facilities are also vampires, and they are draining the aquifers dry. U.S. data centers consumed more than 200 billion gallons of water last year, with roughly 93 percent of that footprint hidden inside off-site power plants where nobody’s looking. In drought-stricken Texas, consumption is projected to explode 714 percent by 2030, yet a state legislative hearing revealed that a staggering 83 percent of the state’s data centers simply refused to comply with mandatory water reporting. This stops being a negotiation and becomes the outright seizure of a shared public resource, with the state acting as armed lookout for the plunder.

It becomes something far more sinister when you remember what a lot of this infrastructure is actually for. Data centers themselves are neutral, but their deployment by the state is not. When the Department of Defense leases military land to build data centers, it isn’t for hosting cat videos. It is the physical construction of a turnkey surveillance apparatus, built on the one category of American ground that is already unambiguously controlled by the government, with none of the normal local checks that even a private developer would face. We are being told to subsidize the very infrastructure that is increasingly being used to watch our every move — and to do it on land we collectively own.

Keep reading

Waymo Robotaxi Crash Rate 68% Lower Than Human Drivers, Study Finds

Though they’ve have racked up quite a few troubling anecdotes on America’s roads, Waymo robotaxis have achieved a crash rate that’s far lower than what’s observed when humans are driving, according to a recent study published by the Insurance Institute for Highway Safety. A private scientific organization funded by insurance companies, IIHS is regarded by many as a superior font of accurate scientific conclusions compared to the government-run and lobbyist-vulnerable National Highway Transportation Safety Administration. 

According to the July IIHS study, “Rise of the Machines: Crash Experiences of Highly Automated Vehicles and Human Drivers,” Waymo vehicles in autonomous mode posted “police-reportable crash involvement rates” fully 68% lower than human-operated vehicles in the same areas and years. Researchers studied crash rates in Austin, Los Angeles, Phoenix and San Francisco. Importantly, the IIHS studied data from 2021 to 2024; to the extent Waymo has improved its programming, the current relative performance may be even better. 

The Waymo edge was highest in Phoenix and Los Angeles (76% and 71% lower crash rates, respectively). San Francisco’s Waymos posted a 35% lower crash rate. In a bit of a headscratcher in terms of the being a huge outlier, Waymo had a 4% higher crash rate in Austin. The study’s authors note that the sample size in Austin was smaller for both Waymo and human drivers. 

Keep reading

Amish farmers fined for not using COVID-era ArriveCan app as legal battle over religious freedom continues

Picture rolling fields, hand-built barns and a way of life largely untouched by modernity. Around two and a half hours west of Toronto, Amish communities continue to farm using animal and human power, rejecting electricity, automobiles and most modern technology.

For the Amish, this is not simply a lifestyle choice. It is a religious practice rooted in their Christian faith. Their communities live much as they have for generations, keeping a low profile, avoiding conflict and turning to a simpler way of life.

That changed during the COVID pandemic hysteria.

Amish families regularly crossed the Canada-U.S. border for family, religious and everyday reasons. When returning to Canada, they were asked the same question as other travellers: had they downloaded the ArriveCan app?

For the Amish, the question created an impossible situation. They do not use smartphones, and their faith prevents them from adopting much of the modern technology required by the app.

Without realising it, Amish families were issued fines that eventually totalled hundreds of thousands of dollars across the community. Some only discovered what had happened when liens were placed on their farms, threatening properties that had been passed down through generations.

Around 40 farms were affected, leaving Amish families facing a legal battle they never expected to fight.

The Democracy Fund stepped in to challenge the cases, with lawyers Marc Joseph and Adam Blake-Gallipeau working to reopen the convictions. So far, 24 families have had their cases reopened and stayed, removing the immediate legal threat. However, eight families were denied that relief and are continuing their appeals through the Ontario courts.

The cases have highlighted the challenges faced by a community that avoids conflict and does not use modern communication systems, yet found itself navigating a complex legal process involving courts, paperwork and technology.

The legal fight continues, with the remaining families seeking to have their cases reconsidered through the appeals process.

Keep reading

Wichita Man Sues City To Rip Out Its 200-Camera Flock Dragnet

A Wichita machinist, Mason Grimmett, sued his own city this week, arguing that the roughly 200 Flock license-plate readers the police department has bolted around town amount to warrantless dragnet surveillance forbidden by the Kansas Constitution. The complaint, filed in Sedgwick County District Court by the Kansas Justice Institute, does not ask for money — rather, it asks a judge to declare the program unconstitutional and shut it down.

The cameras are made by Flock Safety, one of the largest plate-reader vendors in the country, and Wichita started installing them in 2020. The institute now describes the city as saturated with them — close to 200 AI-powered cameras photographing every passing vehicle and logging its make, model, color, plate, and identifying marks into a searchable database. Grimmett, a lifelong Wichita resident who testified against the program at a public meeting in June, argues that the system violates Section 15 of the state constitution, which guards against unreasonable searches and seizures, and that the city is, in the complaint’s language, tracking and databasing law-abiding Kansans “without a warrant, without probable cause, without reasonable suspicion” and without any judicial oversight at all.

The Cops Who Already Abused It

Grimmett points to several local cases in which Kansas police used Flock cameras to stalk people, including abuse by former Kechi Police Lt. Victor Heiar, who was sentenced to eighteen months of probation for using Flock cameras to track a woman. According to local reports, a former detective was also accused of using plate-reader technology to stalk his wife, and a Sedgwick-area police chief reportedly ran an ex-girlfriend through the cameras 164 times. The institute is not disputing that the cameras help solve crimes. Its point is that they photograph everyone, all the time, and that the same access which cracks a burglary case also lets a jealous cop follow whoever he pleases.

“This is America, and we don’t tolerate warrantless surveillance,” said Sam MacRoberts, the institute’s litigation director, who says that if the government wants to monitor a citizen it ought to go to a judge and explain why. Grimmett says he cannot drive to work, the gym, a doctor’s appointment, a polling place, or a meeting of the local Sunflower Privacy Alliance without being catalogued by the department and by Flock.

Keep reading