The funding document accuses CISA of having been “more focused on censorship than on protecting the Nation’s critical systems.” The budget seeks to eliminate offices that were “used as a key hub in the Censorship Industrial Complex” and to “target the President.” CISA has already been hollowed out. Jen Easterly, the last Senate-confirmed head of the agency, who departed when President Donald Trump returned, said at a conference that staff members were being culled with a “mandate for loyalty to a person over loyalty to the Constitution of the United States of America.” The agency has been without a Senate-confirmed leader since Trump took office.

Altman's hard-charging approach is a poor fit for a technology that stands to cause so much disruption. Core to the complaints of his critics is that Altman's competitive impulses have pushed safety concerns to the sidelines. Promises to generously support "superalignment" teams -- those tasked to make sure the AI acts in a way that's complimentary to humanity's aims — turned out to be hollow. A white paper published by the company on Monday, alongside The New Yorker article, seemed timed to counteract these fears, but it won't come close to settling them.

Sometimes it’s possible to agree on a goal but disagree on how to get there. Such is the case with the Federal Communications Commission’s recent surprise ban on all foreign-made home routers — a blow using a blunt instrument that will inflict unnecessary trauma on the industry and most likely lead to higher prices for consumers.

It’s too early to say whether these cases are, in Lanier’s words outside court, “a referendum — from a jury to an entire industry — that accountability has arrived.” But if we can’t yet say it has arrived, it is certainly banging hard on the door. For the first time, these cases have leveled the playing field, one that the lobbyists who have scuttled hopes of effective policymaking could not tilt.

The White House has concepts of a plan on artificial intelligence. They were unveiled on Friday in the administration’s National Policy Framework for AI, a “road map” to legislation, Republican House leaders said, designed to help the US “beat” China and help protect children. In reality, it’s a blueprint for AI companies to carry on with business as usual.

In the debate about the military’s use of artificial intelligence, prompted by Anthropic’s dispute with the Pentagon that’s now headed to the courts, much has been said about the concerns related to autonomous killing. Less examined has been the AI company’s second point of contention: how AI might be used to conduct mass surveillance on Americans. A recent study offers a glimpse at the root of the company’s well-justified trepidation.

Apple’s strategy seems quite straightforward: A cheaper laptop puts it in direct competition with Google’s Chromebook, now a mainstay of classrooms globally, plus a whole universe of Windows-powered laptops. It could lock in the next generation of Apple laptop users. Yet Apple could have justifiably made that kind of move at any point in the Chromebook’s almost 15 years on the market. Instead, Apple stayed firm to its premium reputation and price point. A confluence of reasons made 2026 the right time for a cheaper MacBook.

On Monday, OpenAI Chief Executive Officer Sam Altman, facing pressure internally and out for the deal he reached with the Pentagon to supplant rival Anthropic, conceded he may have rushed into the agreement and was looking to make some amendments. In a memo to staff that Altman subsequently shared on X, his half-baked thinking was on full display. Reading between the lines, Altman appears to be aware he made a severe miscalculation.

After announcing a sudden and surprising deal with the Pentagon, OpenAI co-founder Sam Altman moved quickly this weekend to show he had won meaningful concessions from military leaders. He hadn’t. Whether through naivete or disregard, Altman has quickly leaped over the ethical and practical red lines that his company’s rival, Anthropic, wasn’t prepared to cross.

Preempted by its customers, Nvidia thus needed its own fresh good news to trump what investors already knew. Hello, margins. Adjusted gross margin in the November-January period was 75.2%, the highest it has been since the second half of 2024. The company forecasts that number to be roughly the same in the current quarter. What’s unclear is just how long Nvidia can maintain this extraordinary profitability as the AI landscape matures.

From several angles, pressure is being applied on Anthropic to fall in line. In Tuesday’s meeting, Amodei must state it plainly: It is not “woke” to want to avoid accidentally killing innocent people. This isn’t a case of an arms maker dictating how the Pentagon must use a weapon it has purchased or against which target. No, this is a responsible company making sure a tool bought for one purpose won’t be recklessly used for another.

Will OpenAI be good for the money? There’s no telling what condition the ChatGPT-maker will be in, as it hurries to monetize away its heavy losses with experiments in ads and shopping, by the time the capacity is ready to be used.

Executives will probably also argue that limits to the technology can reduce the possibility of abuse. Don’t believe them. One idea being discussed, according to the Times, is that the glasses might recognize only people you are connected with, something that seems vaguely pointless — don’t you know who your friends are? The second, that it would work only on people with a public profile on Facebook or Instagram, is a safety disaster: What women might wish to disclose publicly to an online audience is vastly different from what they may wish to share with a man sitting opposite them on a bus.

All of this has given Pichai just enough latitude to make his pitch, which is: We’re seeing results and now it’s really time to go gung-ho. At the rate things are going, this time next year $185 billion might look quaint.

It’s easy to get carried away: When the bots start to talk as if they’re planning to take over the world, it can be tempting to take their word for it. But the world’s best Elvis impersonator will never be Elvis. What’s really happening is a kind of performance art in which the bots are acting out scenarios present in their training data. The more practical concern to have is that the powers of autonomy the bots already have is enough to do significant damage if left untethered. For that reason, Moltbook and OpenClaw are best avoided for all but the most risk-tolerant early adopters.