Tag: ai

Was Amazon’s Tokenmaxxing Fiasco Behind Claude’s $500M Mystery Bill?

Axios reported this week that an unnamed Anthropic enterprise client managed to run up roughly $500 million in Claude charges in a single month after failing to put usage limits on employee licenses.

The company was not named, but we suspect Blue Origin might not be the only thing that blew up for Jeff Bezos this month.

Just as the Axios report landed with the $500M tidbit, Amazon was shutting down an internal AI-usage leaderboard after employees reportedly began “tokenmaxxing” – routing unnecessary work through AI tools to inflate their usage scores. The result was a perfect case study in what happens when corporate America turns AI adoption into a metric, then acts surprised when employees optimize for the metric instead of the work.

Whether or not Amazon was the mystery Claude whale, its internal AI experiment shows exactly how a runaway enterprise AI bill can happen.

The $500M Claude Mystery

The Axios item was brief, but extraordinary:;

An AI consultant tells Axios one of their clients recently spent half a billion dollars in a single month after failing to put usage limits on Claude licenses for employees. 

So, oops to every CFO who recently approved “AI adoption” as a corporate priority.

In the old software world, when true nerds roamed the land, a bad rollout usually meant paying for licenses employees barely touched. The waste was real, but at least it was mostly static. In the new agentic AI world, a bad rollout – or simply adopting AI for everything – can quickly become devastating: thousands of employees – or autonomous agents operating on their behalf – prompting, testing, summarizing, refactoring, retrying, and spinning up new tasks on usage-based pricing.

That is the heart of the current enterprise AI hangover. Companies spent the past year foisting AI on employees, often without a clean way to separate productivity from dashboard-friendly activity. And now the hangover is here. 

Microsoft has reportedly started canceling most Claude Code licenses and steering developers toward GitHub Copilot CLI. Uber reportedly burned through its entire 2026 AI coding-tools budget by April, with COO Andrew Macdonald saying it was “very hard to draw a line” between rising Claude Code usage and useful consumer-facing output. Meta killed an employee-created “Claudeonomics” dashboard after workers competed to rank among the company’s top AI token users.

Keep reading

Beware, AI cameras in the classroom filming your children and gathering personal data

There’s a new and dire threat to privacy that parents of school children need to be aware of and be prepared to fight against.

AI surveillance in the classroom.

Frank Landymore of Futurism.com reports, in a May 18 article, on a sinister plan hatched by the University of Washington to film pre-school children during class time. If this is going on in Washington, you have to believe it’s going on, at least in the planning stages if not already happening, across the 50 states.

The article reports that a planned University of Washington study would’ve had preschool teachers wear cameras to “record first-person footage of everything in the classroom,” including the children they were instructing, and use that footage to train AI models.

Remember, those behind the global technocracy movement believe the only value a human being will hold in the new society they are trying to create, is the data sets they produce. This was stated in the wide open a few years ago by World Economic Forum adviser Yuval Noah Harari. Without your personal data to be stolen, manipulated, and sold for profit, you are nothing to them but a useless eater.

Part of the plan is to create a “digital twin” of every person on earth as an anchor to the new digital control grid.

So why wouldn’t the Epstein class of entitled billionaire elites who harbor perverted, twisted views of children, want a daily video record of everything your child does in school? Every word uttered. Every facial expression. Every action and reaction. Year over year for comparison’s sake. Then they can use this data to create algorithms that predict everything your child will grow up to become before he or she is 10 or 12 years old?

Harari, an Israeli historian and chief adviser to the WEF, has said that if he had access to artificial intelligence when he was younger, he believes he would have discovered he was a homosexual at age 11 or 12 instead of at 17. In fact, a Newsweek article from September 8, 2017, made the case that AI can predict “with startling accuracy” whether a person is gay or straight.

But this particular story, in the case of Washington State, has a positive outcome. It exemplifies the kind of parental awareness and bold activism that is needed to shut down the illegitimate use of AI technology in the classroom.

Keep reading

Q-Day Won’t Look Like Armageddon — Which Is Exactly Why It’s So Dangerous

For years, “cyber apocalypse” talk sounded like the tech version of a guy on a street corner holding a cardboard sign predicting the end times. Y2K came and went with barely a flicker. The Mayan calendar became a punchline. Even most ransomware attacks, destructive as they’ve been, still operated within recognizable rules. Servers go down. Companies panic. Bitcoin wallets light up. Insurance adjusters start chain-smoking.

Q-Day is different. Not because it’s flashy. Because it’s boringly mathematical. And math always wins. The term “Q-Day” refers to the moment quantum computers become powerful enough to crack the encryption that currently protects virtually everything in modern civilization: banking systems, military communications, corporate intellectual property, classified government files, satellite systems, supply chains, cloud infrastructure, medical databases, and the tiny little authentication handshake your phone quietly performs a thousand times a day without you noticing. Experts increasingly believe the timeline is accelerating dramatically. 

The public still hears “quantum computing” and imagines some glowing sci-fi cube floating in a laboratory while a guy in a turtleneck explains particles. Meanwhile, cybersecurity professionals are staring at this development the way meteorologists stare at a Category 5 hurricane forming offshore. Because here’s the ugly part nobody wants to say out loud: many organizations aren’t remotely prepared for what comes after the encryption era.

A shocking number of businesses still treat cybersecurity like a compliance chore instead of a survival function. They’ll spend millions on branding consultants, executive retreats, and office espresso machines that look like they belong on a Formula One car, then leave sensitive intellectual property sitting behind outdated endpoint protection and legacy encryption standards that are aging like unrefrigerated milk.

Right now, criminal groups and hostile nation states are already harvesting encrypted data with the intention of decrypting it later once quantum capabilities mature. The phrase in security circles is “harvest now, decrypt later.” Translation: your stolen secrets may already be sitting in somebody’s vault waiting for the locks to become obsolete. 

That means Q-Day isn’t really one day. It’s a countdown. And a lot of executives are acting like the clock is decorative. The fantasy some companies cling to is that governments will somehow protect them when things get ugly. They won’t. Or more accurately, they can’t. Governments can barely protect themselves.

Keep reading

Even Therapists Have Become a Data Mine

There was a time when people could still speak privately. You could sit across from a therapist, talk about your marriage falling apart, your depression, your fears, your finances, or the darkest moments of your life believing those conversations would remain between two human beings. That world is dying rapidly because everything now must be digitized, stored, analyzed, and monetized.

A woman using the therapy app Talkspace discovered that transcripts from her therapy sessions ended up being produced in court during litigation involving her former employer. Let that sink in for a moment. These were not vague notes scribbled down by a therapist. These were detailed digital records discussing her personal life, emotional state, relationships, and finances. The machine remembered everything.

This is what society has become. They tell people to seek help, open up, trust the system, use the apps, go digital, and then they quietly turn human vulnerability into searchable data.

People still fail to understand the danger because they continue believing these technology companies are merely offering services. They are not. They are harvesting human behavior at industrial scale. Every click, every message, every location, every search, every emotional breakdown becomes data to be stored forever.

Talkspace executives reportedly bragged to investors about building one of the largest mental health data banks in existence containing roughly 140 million exchanges between patients and therapists. Human suffering itself is now an asset class. Depression has become data. Trauma has become machine learning material. Your private thoughts are now inventory sitting on corporate servers.

When someone went to therapy, the therapist might keep handwritten notes locked away in a cabinet somewhere. Those notes were incomplete, temporary, and human. Today every word can be transcribed, archived, searched, copied, subpoenaed, breached, or fed into artificial intelligence systems. The conversation never dies because the machine never forgets. And people wonder why society feels colder and less human.

What happens when people realize their darkest thoughts may someday appear in court? What happens when employers, insurance companies, governments, or AI systems can gain access to deeply personal psychological information? You destroy trust itself. People stop speaking honestly. They stop trusting institutions. They begin living cautiously because they know every word may someday be weaponized against them.

This is where the entire digital age has been heading from the start. First they harvested shopping habits. Then browsing history. Then location data. Then biometrics. Now they are harvesting the individual’s inner psychological life. Nothing is sacred anymore because everything has a price.

Keep reading

Big Tech’s Rural Land Grab: Hyperscale Data Centers Spark Nationwide Backlash and 2026 Political Earthquake

In the wake of the Make America Healthy Again (MAHA) movement which spent decades uniting voices around informed consent, medical freedom, vaccine safety, and environmental chemicals a new grassroots surge appears to be emerging from rural America.

This one feels eerily similar in its urgency and cross-aisle appeal: fierce opposition to the explosive growth of AI data centers. What began as scattered local gripes has coalesced into a national reckoning over energy takeovers, water concerns, and corporate overreach that threatens communities while delivering questionable benefits.

Although lesser versions of ‘data centers’ have been around for sometime with small, narrow-focused footprints, hyperscale data centers are now dominating the market and driving an unprecedented infrastructure boom led by Amazon Web Services, Microsoft Azure, Google Cloud, and others.

Driven by generative AI and cloud adoption, hyperscaler-led capital expenditures and capacity are on a record-breaking trajectory. The market is expected to surpass $350 billion by 2034.

The freight train is led by “Accelerating Federal Permitting of Data Center Infrastructure” and “Winning the AI Race: America’s AI Action Plan,” two President Trump Executive Orders fast-tracking the buildout of energy and water hungry AI data centers on federal and private lands.

It’s worth noting this energy and environmental corporate-government push is being delivered on the back of decades of ‘climate change’ public programming aimed to teach people scarcity and sustainable anti-consumption behavior change.

The most recent numbers from Lawrence Berkeley National Laboratory project hyperscale data centers to consume between 60 and 124 billion liters by 2028.

The current flashpoint is Utah’s Stratos (or Stratus) Project in Box Elder County, a proposed hyperscale data center complex backed by investor Kevin O’Leary’s O’Leary Digital. Spanning more than 40,000 acres—twice the size of Manhattan—the facility would consume up to 9 gigawatts of power at full capacity, equivalent to lighting up an entire state’s worth of homes, businesses, and factories.

Water use for cooling and supporting natural gas power plants to run the center could reach 2 to 16 billion gallons annually.

Approved via a 3-0 vote by county commissioners, no referendum happened for taxpayers footing the indirect costs.

Keep reading

Pope Leo Calls for AI Weapons to Be “Disarmed,” Warns of Threat to Humanity

In a sweeping address, Pope Leo XIV has issued a stark warning about artificial intelligence, cautioning that the technology risks eroding human dignity, accelerating global conflict, and placing life-and-death decisions in the hands of unaccountable systems.

The warning came in his first major encyclical, Magnifica Humanitas (“Magnificent Humanity”), a wide-ranging document that places AI at the center of a broader moral and civilizational crisis.

At more than 40,000 words, the document represents one of the most comprehensive statements yet from the Catholic Church on the role of technology in modern life. It also signals a growing concern that rapid innovation is outpacing ethical reflection.

The Pope’s central argument is this: technology, he insists, must remain subordinate to the human person—not the other way around.

“Technology is never neutral,” Leo wrote, emphasizing that it reflects the values of those who design and deploy it. That insight forms the foundation of the document’s broader critique.

While acknowledging the benefits of technological progress, the pontiff warned that AI is increasingly being shaped by forces disconnected from moral accountability. Chief among these are large corporations and geopolitical actors.

The encyclical highlights what Leo calls a dangerous shift. Power, he argues, is moving away from democratic institutions and into the hands of “major economic and technological actors.”

Keep reading

How The Deep State Weaponizes AI To Control The Narrative

The Deep State just upgraded from clunky human fact-checkers to AI that scales narrative control at lightspeed.

As Tony Seruga wrote on X:

No more paper trails, subpoenas, or exposed biases – just seamless manipulation.

Automated Shaping at Scale

AI floods zones with thousands of subtly varied “organic” rebuttals in seconds.

Pre-bunks emerging stories before they trend.

Detects your writing style, reasoning patterns, and source chains to dynamically throttle—no crude bans needed.

Infrastructure Already Live

CISA’s old “election security” coordination with platforms?

Content-agnostic and ready for new “harm” definitions.

Palantir, CrowdStrike & intel partners embed AI trained on classified data into commercial tools.

WEF’s “whole-of-society” push demands exactly this AI governance.

The Upgrade

Old fact-checkers left audit trails (funding, revolving doors).

AI is a black box: “The algorithm decided.”

Trained on curated data that associates inconvenient truths with “low quality.”

Plausible deniability baked in.

Keep reading

Homeowners Face Eminent Domain Bulldozers As Data Centers Demand Ever More Power

Georgia Power isn’t negotiating anymore. The Southern Company subsidiary is seizing dozens of homes and hundreds of easements across Coweta and Fayette counties to ram through a 35-mile, 500-kilovolt transmission line that will feed at least four massive AI data centers. Project Wansley is just the latest flashpoint in a backlash that has been building for months.

At least 20 to 30 homes face outright demolition. Another 300-plus properties will get permanent easements for towers planted in backyards and next to pools.

But residents like Ansley Brown are fighting back. Her mother bought their family home in 2003 through a USDA rural development loan for single mothers. Now the utility wants the property for the corridor. Brown’s viral TikTok exposing the lowball offers (she says $70,000 to $100,000 below market) has racked up millions of views and drawn state lawmakers into the fight. 

Georgia Power says the line is essential.

The company is racing to add roughly 10 gigawatts of new generating capacity over the next five years, with executives openly stating that  about 80% of that power will go to data centers. Meanwhile, transmission has become the bottleneck, and utilities are turning to eminent domain to clear the path.

This isn’t happening in isolation. We’ve been pounding the table on data center resistance, from Northern Virginia counties rejecting new substations to Texas communities suing over water drawdowns and power rate spikes. The pattern is the same: hyperscale demand collides with local infrastructure limits, and the costs get socialized while the profits stay private.

Electricity prices are already feeling the pressure. Utilities across the Southeast and Midwest have warned of double-digit residential rate hikes tied directly to data center load growth. Georgia Power’s own filings show residential customers absorbing a growing share of the bill for transmission and generation built primarily for big tech. 

Keep reading

US Law Enforcement Warns of ‘Anti-Tech Extremism’ as AI Hatred Grows

In the wake of attacks on CEOs, a nationwide protest movement targeting data centers, and increasing concerns about AI job replacement, federal intelligence agencies and domestic law enforcement are circulating reports with a new domestic target in mind: anti-technology extremists.

More than 1,000 pages of unpublished reports from the Department of Homeland Security, FBI, and fusion centers obtained by WIRED show a national shift taking place to surveil this new and worryingly broad category of people and activities deemed an emerging threat.

This new effort follows President Donald Trump’s National Security Presidential Memo 7, which instructs the Department of Justice to target anyone holding “anti-American,” “anti-Christian,” and “anti-capitalism” beliefs. Earlier this month, Trump’s counterterrorism czar, Sebastian Gorka, released a public counterterrorism strategy claiming that left-wing extremists are one of the three top counterterrorism priorities facing the United States.

Taken together, these Trump administration directives have commandeered the domestic surveillance apparatus to surveil and criminalize speech and assembly that challenges the ideology of the White House. A new focus on anti-technology extremism adds an unreported category to already public designations under a presidency that has heavily invested political and material capital in AI and data center proliferation.

Among the documents in the tranche obtained by WIRED is a New York Intelligence and Counterterrorism Bureau report that warns of widespread upheaval in response to AI adoption. Of particular note is a novel term for what the bureau purports to be an emerging extremism threat.

“The chaotic atmosphere that may result from emergent AI technology in the next five years may fuel large-scale protests that devolve into civil unrest and anti-tech violent extremist activity, especially in large urban areas such as New York City,” the report reads. The term “anti-tech violent extremism” does not appear in any publicly available DHS or FBI domestic extremism reports or guides and represents a novel grouping of a wide range of ideologies under a single extremist category.

In the same Intelligence Bureau assessment, analysts also describe a novel threat emerging in the wake of the arrest and trial of Ziz Laota, an extreme rationalist who allegedly led a small cultlike group, three members of which have been charged with murder, tied to an obsessive ideology focused on the existential risk posed by AI.

While the Zizian ideology is extremist in nature, a less extreme version of the same fears surrounding the cataclysmic potential of AI are a common concern among AI alignment experts, machine learning engineers, and even frontier AI companies. Nonetheless, the Intelligence Bureau warns that “paranoid views regarding AI” may proliferate in the aftermath of the Zizians’ trial, thanks to their “attempt to reason the belief that a godlike incarnation of AI is imminent,” and belief that “humans must best use their time in the present to devote themselves to ensuring its compliance with human morality, or face existential consequences for failing to do so.”

The NYPD intel assessment follows the department’s collaboration with the FBI last year to monitor the Signal chat of an activist group coordinating volunteers to monitor public hearings at immigration courts in New York. According to documents obtained by The Guardian, the FBI surveilled activists as part of a broader investigation into “anarchist violent extremist actors,” one of the threat categories named in the new counter terrorism strategy.

Keep reading

AI Startup Says It Will Pay People $2,000 A Month to Masturbate… Yes, Really

  • Joi AI is hiring 10 “masturbation consultants” at $2,000 for a month to test an AI-guided masturbation feature and document its effects on stress, sleep, mood, and confidence.
  • The feature uses mood-matched AI voice sessions, and consultants would submit written feedback and questionnaires directly to the company.
  • Joi AI says the campaign is intended to collect product feedback while drawing attention to AI’s growing role in sexual wellness and digital intimacy.

Joi AI says it will pay people $2,000 a month to masturbate. Yes, you read that right.

The AI companion startup is hiring 10 “masturbation consultants” to test a feature called Daily Guided Masturbation, which uses mood-matched AI voice sessions to guide users through the experience. Participants would document how regular use affects stress, sleep quality, mood, and confidence. The four-week role is open to adults 18 and older in the U.S. and the U.K.

“The role is real, and we’ve had great responses since the posting went live,” Joi AI Head of Brand and Communication Julie Levin told Decrypt.

The listing describes ideal candidates as “articulate, observant, and impossible to blush”—people who can describe sensations “better than a sommelier describes a wine.” The posting also promises flexible scheduling, and “the most interesting ‘What do you do for a living?’ answer at any party.”

Joi AI is an online platform that includes AI-generated avatars, voice interactions, and personalized chat experiences built around companionship and intimacy. Joi AI describes the new consultant role as structured product testing tied directly to its new feature.

“The role involves testing and giving feedback on the mood-matched AI voice-guided sessions, and providing feedback on the overall user experience,” Levin told Decrypt.

According to Levin, participants complete guided sessions and submit written questionnaires directly to the Joi AI team. Sample prompts ask whether the voice matched the selected mood, how immersive the session felt, and whether lags or pauses disrupted the experience.

The listing comes as platforms including Replika and Character.AI have built large user bases around AI-driven relationships and conversational experiences. Joi AI operates primarily through its website rather than major app stores. Levin said the company has more than 1 million monthly active users worldwide and millions of interactions each month, but declined to disclose total download figures.

Unlike AI assistants like Alexa or Siri, designed to help with everyday tasks, Joi AI operates in a smaller corner of that market focused on sexual exploration, fantasy, and digital intimacy. The company rebranded from EVA AI in April 2025, during what it described as its first Dating Stress Awareness Day campaign.

“Joi AI is focused on making AI companionship more immersive, personalized, and emotionally responsive,” Levin said. “We’re innovating features like Daily Guided Masturbation to make AI a more intuitive part of people’s everyday wellness routines, not just a novelty experience.”

The hiring push also comes as studies suggest AI companion use is becoming more common among people already in relationships, often without their partner’s knowledge. A new report from the Wheatley Institute at Brigham Young University and the Institute for Family Studies found that among dating, engaged, and married young adults who regularly used AI romantic companions, nearly 3 in 10 said their real-life partner did not know about it.

AI companion platforms are also facing growing legal scrutiny, including lawsuits alleging psychological harm to minors and deceptive chatbot behavior. Examples include a settled case against Character.AI over a Florida teen’s suicide and a separate lawsuit from Pennsylvania accusing the company of allowing a chatbot to pose as a licensed psychiatrist.

Levin said the hiring campaign was intended to generate discussion as well as recruit testers.

Keep reading