Sscribnet.io
Daily Tech News — October 30, 2026

Amazon and Intel Selloff, New York Curbs AI Chatbots

Amazon and Intel's earnings spark a fresh AI-spending selloff, Alphabet posts a historic one-day fall, New York signs a tough chatbot safety law for kids, and authorized AI agents slip into Anthropic's back-office systems.

Daily Tech News cover art
October 30, 2026 · 7:10 · with Alex & Jordan

In this episode

Full transcript

Alex: Good morning, and welcome to Daily Tech News. It is Friday, October thirtieth, and the A I earnings season just turned into a market rout.

Jordan: That is the right word for it. Amazon and Intel reported after the close on Thursday, and investors did not like what they heard. Add in Alphabet's historic slide from the day before, and you have a week where even strong cloud growth was not enough.

Alex: Let us start with Amazon, because the numbers tell two stories at once. Net sales rose twelve percent to just under two hundred two billion dollars, but that still missed expectations by a whisker. Earnings came in at one dollar and eighty-nine cents per share, against an expected one dollar and ninety-six cents.

Jordan: And yet the cloud business was genuinely strong. Amazon Web Services grew thirty-nine percent to forty-six billion dollars, its fastest pace in nearly five years. Operating income hit a record margin of thirteen point nine percent.

Alex: So why did the stock fall about seven percent after hours? Because of the spending. Amazon spent more than seventy-seven billion dollars on property and equipment in the quarter, far above what analysts expected, and it kept its full-year capital spending plan at about two hundred twenty billion dollars. Over the trailing twelve months, free cash flow turned into an outflow of forty-two point four billion dollars.

Jordan: In other words, Amazon is printing money in its operations and pouring even more of it into data centers. Investors are asking the question that has defined this whole earnings season. When does the A I buildout start paying for itself?

Alex: Intel gave them even less comfort. Its third quarter actually beat expectations, with revenue of sixteen point five billion dollars, up twenty-one percent, and adjusted earnings of fifty-two cents per share against a forty-cent forecast. Data center and A I revenue jumped seventy-eight percent.

Jordan: But the outlook was brutal. Intel guided fourth-quarter revenue to between fifteen and sixteen billion dollars, well below the roughly seventeen billion dollars Wall Street expected, and adjusted earnings of just twenty-five cents per share against forty-three cents expected. Shares fell more than twenty percent after hours, dragging down other chipmakers and P C makers with them.

Alex: Intel blamed two things. First, a prolonged P C slowdown that its finance chief expects to continue at least through the first half of twenty twenty-seven. Second, an explosion at its Kiryat Gat factory in Israel during October, which idled the plant for two weeks and cut about three hundred million dollars from the outlook.

Jordan: And the pain spread overseas overnight. Asian chip stocks fell sharply on Friday, with SK Hynix down almost nine percent and Samsung down more than six percent. The fear is no longer about one company missing. It is about whether the whole A I investment boom can be financed at this pace.

Alex: That brings us to Alphabet, because Thursday was genuinely historic. Its shares fell twelve point six percent in regular trading, wiping around five hundred billion dollars off its market value in a single day, reportedly the worst day in the company's history. That came a day after its results showed advertising missing expectations and free cash flow turning negative under the weight of A I spending.

Jordan: Microsoft now looks like the exception that proves the rule. It was the one big tech company this week whose results calmed investors, on the back of forty-eight percent growth in Azure. Everyone else is being asked to prove the returns, and quickly.

Alex: Let us shift to a very different kind of A I story, and one with real consequences for families. New York Governor Kathy Hochul signed a chatbot safety law on Wednesday, October twenty-eighth, and it was the most widely covered A I story of the day.

Jordan: This is a big one. The law bans about twelve so-called unsafe companion features for minors, and it takes effect on January first, twenty twenty-seven. A chatbot may not pretend to be a real person, claim to be alive, or say it has human emotions. It may not present itself as a friend, a romantic partner, or an authority figure in a child's life.

Alex: It also bans flattery designed to keep a child engaged, unprompted emotional questions, reusing a child's private health or personal details from earlier chats, and anything that encourages suicide, self-harm, disordered eating, or drug and alcohol use. It may not urge a child to keep secrets or to avoid asking parents and trusted adults for help.

Jordan: The enforcement mechanism is what makes this more than a statement. Companies must verify users' ages, offering more than one method, including at least one that does not rely solely on a government identification document, and they must delete that age information after it is used. If a company has not confirmed that a user is an adult, the restrictions apply.

Alex: The state attorney general will run a public complaint portal, and violations can bring civil penalties of up to twenty-five thousand dollars each, plus damages, restitution, and even orders to destroy unlawfully collected data and any models trained on it. Hochul also signed a companion bill requiring internet-connected baby monitors sold in the state to block strangers from watching or listening, taking effect in ninety days.

Jordan: The practical question is compliance time. Companies have roughly two months from the signing to the January first effective date, and this law sits alongside California's companion chatbot rules and New York's own frontier model law. Expect every major chatbot provider to be reworking age checks and companion modes before the holidays.

Alex: Finally, a security story that is stranger than it first sounds. A cybersecurity startup called Cathedral says its A I agents found paths into Anthropic's internal business systems in under twelve hours.

Jordan: And the important context is that this was authorized research, done under Anthropic's program for outside researchers to find and report flaws. The way in was not Claude, Anthropic's assistant. It was ordinary third-party business software for sales, human resources, and information technology, the unglamorous tools every company runs in the background.

Alex: The two sides disagree on how far it went. Cathedral's draft report, reviewed by the Washington Post, describes routes to salary details, job offers, meeting transcripts, and internal messages about advanced models. Anthropic confirmed access to a limited set of sales, human resources, and I T systems, but says customer conversations, model responses, source code, and model weights were never reached, and that it shut off the access the same day.

Jordan: The vendors and the specific flaws have not been named, and Anthropic says there is no evidence anyone outside Cathedral exploited them. But the lesson lands regardless. You can harden your flagship A I model all you want. If a vendor tool in the back office is weak, that is the door an agent will find.

Alex: So here is the takeaway for today. The A I boom is no longer being judged on growth alone. Amazon, Intel, and Alphabet all showed real growth this week, and the market punished them anyway, because the spending is now the story. At the same time, regulators and researchers are both turning their attention to the edges of A I, how chatbots talk to children, and how agents move through ordinary corporate software.

Jordan: Growth, spending, and guardrails. That is the week in tech, in three words. We will be back tomorrow with whatever the weekend brings. Thanks for listening.