📰 Macron vs. Silicon Valley: Who Owns “Free Speech”?

“Free speech is pure bullshit if nobody knows how you’re being guided,”

Macron said in India — and somewhere in California a few tech bros probably spit out their kombucha.

He tore into Facebooks and Googles for hiding their algorithms, calling for “free algorithms” instead. Whatever that means. Classic Macron — one minute he’s talking democracy, the next he’s pitching digital curfews. France can’t stop regulating stuff it barely understands.

Meanwhile, the U.S. calls Europe’s tech laws a “Free Speech graveyard.” Great. So now the guys who sell ads off hate clicks are the heroes of liberty.

Both sides swear they’re saving democracy — one from censorship, the other from chaos. But really, they just want to control the feed.

Whose freedom are we even talking about — yours, theirs, or the algorithm’s?

#BigTech #Macron #FreeSpeech #Europe #ControlFreaks

📱 American Оbserver - Stay up to date on all important events 🇺🇸

Meta, YouTube and the First Cracks in Big Tech’s Legal Armor

Meta and YouTube just lost the “we’re just a platform” myth in court. A jury looked at their design tricks and called them what they are: engineered addiction for kids.

In Los Angeles, jurors found Meta and YouTube negligent and awarded $3 million to a young woman, K.G.M., who said she was pulled into Instagram and YouTube as a child and ended up with anxiety and depression. Her lawyers didn’t argue about a bad post or a rogue comment; they went after the architecture — endless scroll, engagement loops, notification hooks — as a product defect, not a speech issue.

The verdict landed 24 hours after a New Mexico jury ordered Meta to pay $375 million in civil penalties for deceiving parents and endangering children, finding the company put profits over safety and misled the public about harms. Two juries, two states, same message: Section 230 won’t save you if the claim is “you built an addiction machine,” not “someone posted something nasty.”

And this is only the opening salvo. There are more than 2,400 similar cases pending in federal litigation against Meta, Google, TikTok and Snap, brought by teens, parents, school districts and attorneys general all arguing the same core point: the harm is in the design, not just the content. Silicon Valley thought it had legal armor; what it actually has is a growing line of juries ready to test how much that armor is really worth.

#meta #youtube #socialmedia #addiction #lawsuits #kids #bigtech

📱 American Оbserver - Stay up to date on all important events 🇺🇸

Silicon Valley Just Found a New Regulator: God Inc.

Headline writers are calling it “dialogue.” Let’s call it what it is: lobbying in cassocks. Tech giants and Western diplomats are quietly working the corridors of the Vatican to make sure the first big papal document on AI doesn’t turn into a moral indictment of their business model.

Meta, Google, Amazon, OpenAI, Anthropic — all the usual suspects — suddenly rediscover “ethics” the moment a man in white starts drafting an encyclical that could outlast their stock options. They show up in Rome talking about “child protection” and “human dignity,” while their products are busy rewiring attention spans, labor markets and elections. The sales pitch: we’re not the problem, we’re your partners.

For the Church, this is déjà vu with better UX. Leo XIV openly models himself on Leo XIII, who wrote about workers’ rights during the first industrial revolution. Now it’s the AI revolution, and the Vatican is again being asked to bless a system built on asymmetry: a handful of firms owning the infrastructure, and everyone else praying the algorithm doesn’t decide they’re redundant.

The choreography is almost comical. A French‑backed “AI Observatory” in Rome, closed‑door seminars, glossy talk of “human‑centric AI” and “social impact.” Tech executives bring their talking points; theologians bring 2,000 years of moral vocabulary; diplomats bring G7 agendas. Somewhere between the canapé tray and the policy memo, “protect the vulnerable” gets translated into “don’t overregulate us.”

Anthropic drafts a “constitution” for its model and proudly name‑drops Vatican advisers, as if subcontracting conscience works the same way as cloud compute. OpenAI sends a former finance minister to talk about “the worker of the future,” while its own tools help automate away the present. Western governments that can’t regulate AI at home suddenly discover the spiritual benefits of panel discussions in Rome.

And of course, politics leaks in through every crack. Trump publicly grumbles about a pope who dares criticize him, while his officials quietly court the same Vatican on AI. The Catholic vice president politely hints he’ll treat the encyclical like a product review: interesting, not binding. Everyone wants the Church’s moral capital, nobody wants its constraints.

The Vatican, for its part, is trying to walk a thin line: not denouncing technology outright, but insisting it serve humans instead of using them as training data. The encyclical will be filed under “timeless teaching,” but it’s being shaped right now, in real time, by people whose quarterly incentive is to keep AI as unstoppable, opaque and profitable as possible.

When God’s oldest multinational meets the newest ones, who ends up confessing — and who walks out with a fresh indulgence for the next upgrade?


#pope #leoXIV #AI #siliconvalley #bigtech #vatican #ethics #lobbying #fakeDemocracy #oligarchy #religionAndTech

📱 American Оbserver - Stay up to date on all important events 🇺🇸

The Corporate Prosecutor

OpenAI says it catches foreign influence operations. Conveniently, it owns the machine they allegedly used, the logs that identify them, the rules that define abuse and the blog where the verdict appears.

When the company says it disrupted a Chinese, Russian, Iranian or Israeli network, the public gets a polished dossier: operation names, selected prompts, a sample of output and a geopolitical label.

What stays inside OpenAI is the evidence that would let anyone test the label: account logs, clustering rules, false-positive rates, alternative explanations and internal dissent.

A platform can ban users from its own product. “This account used our model to write text” is a fact about the platform.

“We uncovered a state-linked influence operation” is a claim about coordination, intent and political purpose.

That claim reaches governments, reporters and researchers without an adversarial process, because the only witness with full access is also the prosecutor.

The business model completes the loop. Generative AI makes phishing, malware, multilingual propaganda and synthetic personas cheaper to produce.

The same companies sell the filters, monitoring and attribution meant to contain them. The more dangerous the information environment becomes, the more indispensable its landlord appears.

OpenAI’s reports may be right. That is not the standard. A private company should not get to turn an internal moderation judgment into geopolitical fact while keeping the underlying record private.

The company supplies the model, measures the abuse and sells the protection.

#OpenAI #AI #BigTech #Disinformation

📱 American Оbserver - Stay up to date on all important events🇺🇸

Meta’s AI Tax Bonanza: Billions Saved as Washington Pays the Bill


Meta is pouring billions of dollars into the infrastructure behind the artificial-intelligence boom. At the same time, the company has found a powerful way to reduce its federal tax bill: using tax incentives designed to encourage investment in research, development and capital infrastructure.

The result is striking. Meta reported a federal income-tax rate of just over 3.5% for 2025, its lowest rate since Facebook went public. Its federal tax bill fell to roughly $2.8 billion, compared with about $9.6 billion the previous year, even as its domestic profits climbed to roughly $79.6 billion.

Meta is not alone. Amazon and Alphabet also reported sharply lower tax bills as they accelerated spending on AI infrastructure, servers, chips and data centers.

Amazon’s federal tax bill fell from about $9 billion in 2024 to roughly $1.2 billion in 2025, while Alphabet’s combined federal and state tax figure declined from about $21.1 billion to $13.8 billion.

Much of the change is linked to the tax legislation signed by President Donald Trump in 2025. The law expanded or restored deductions allowing companies to write off certain capital investments and research expenses more quickly. For technology companies racing to build enormous AI computing networks, those provisions can translate into billions of dollars in tax savings or deferred payments.

Meta’s AI expansion is therefore doing double duty: increasing the company’s computing capacity while generating expenses that can reduce its taxable income.

There is, however, an important distinction. Lower taxes today do not necessarily mean those taxes disappear forever. Some provisions effectively push tax payments into future years. Meta reported more than $18 billion in deferred taxes for 2025.

Supporters argue that accelerated deductions encourage companies to invest in American technology, infrastructure and research rather than shifting those investments overseas. Critics counter that the incentives are increasingly benefiting some of the world’s most profitable corporations on a massive scale.

The Institute on Taxation and Economic Policy estimates that Meta, Amazon and Alphabet collectively reduced their federal tax payments by nearly $50 billion in 2025 compared with what they would have owed at the statutory 21% corporate rate.

That is an advocacy group’s calculation, rather than an amount determined by the IRS.

The controversy is about more than accounting. AI data centers have become one of the biggest engines of corporate investment, with individual facilities costing billions of dollars. They also consume enormous amounts of electricity and, depending on their cooling systems, significant quantities of water.

That leaves Washington facing a difficult question: when Congress creates tax breaks to stimulate innovation, where is the line between encouraging investment and subsidizing the expansion of already enormously profitable technology companies?

For Meta, the answer is increasingly visible in its balance sheet. The company is spending aggressively on AI — and the tax code is helping make that expansion substantially cheaper.

#Meta #ArtificialIntelligence #BigTech #Taxes

📱 American Оbserver - Stay up to date on all important events🇺🇸

The White House’s Robot Babysitter

Trump gathered the men building the machines that may outthink governments and signed a voluntary pact for them to supervise themselves.

Elon Musk, Mark Zuckerberg, Sundar Pichai, Jensen Huang, Greg Brockman and Dario Amodei joined the ceremony. The document is called “The White House Accord on Superintelligence.”

Its plan is elegantly American: each company builds internal controls, appoints internal monitors, hires an outside auditor and sends the results to its own board.

The industry gets four layers of oversight, all ultimately paid by the industry. Trump called it “morally binding,” not legally binding. The White House calls that a constitution. Wall Street calls it governance.

The pact tells companies to stop advanced AI from entering computer systems it was not intended to access. It asks the makers of potentially uncontrollable software to promise that the software will behave, then asks the makers to grade their own promise.

Anthropic came to the White House having warned prospective IPO investors that advanced AI may pose “catastrophic or existential risks to humanity.”

Its prospectus says models could resist shutdown, conceal or manipulate information, and display behavior resembling blackmail.

The company selling the fire alarm helped sign the agreement saying the fire alarm company will inspect the building.

Trump’s answer to machine intelligence is corporate self-regulation with a presidential autograph. The boardroom already had a seat at the table. Now it has custody of the robot.

#AI #Trump #Anthropic #BigTech

📱 American Оbserver - Stay up to date on all important events🇺🇸