Tech News — Week of 2026-03-22 to 2026-03-26#

Story of the Week#

Big Tech’s liability shield has officially cracked under the weight of algorithmic addiction. In a landmark legal week, separate juries slammed Meta and Alphabet with multi-million dollar verdicts—including a $375 million penalty in New Mexico and a $6 million ruling in Los Angeles—for intentionally designing addictive products that concealed known harms to children and young adults. By legally equating social media’s infinite scroll to Big Tobacco-style personal injury, these rulings validate the theory of algorithmic negligence and open the floodgates for thousands of lawsuits that could permanently reshape consumer software.

Top Stories#

[OpenAI Kills Sora and Shelves “Adult” Bots] · Ars Technica In a ruthless pivot toward enterprise dominance ahead of an anticipated IPO, OpenAI abruptly shut down its hyped AI video generator Sora, instantly vaporizing a $1 billion licensing deal with Disney. The company also killed plans for a sexualized chatbot over internal fears of users forming unhealthy, sycophantic attachments, proving Sam Altman is clearing the decks of any risky side quests.

[Arm Abandons 35-Year Strategy to Build Its Own AI Chips] · The New York Times Breaking from its historic IP-licensing-only model, Arm is stepping directly into the silicon arena to sell its own physical “AGI CPUs”. Aimed at feeding the insatiable data center demand for power-efficient agentic AI workloads, Meta has already signed on as the debut customer and co-developer for this high-core-count hardware.

[Agentic AI Takes Control of the Desktop] · Engadget The era of AI operating your computer for you is here, with Anthropic upgrading its Claude tools to autonomously navigate macOS web browsers, open files, and execute developer software via simple point-and-click. Meanwhile, Google globally launched Gemini 3.1 Flash Live, delivering sub-300-millisecond conversational latency and real-time vision capabilities to power fluid, human-like voice interactions.

[FCC Drops the Hammer on Foreign-Made Routers] · TechCrunch The US government effectively froze the consumer networking market by banning the import of all new foreign-made Wi-Fi routers due to severe cybersecurity vulnerabilities. The sweeping directive heavily targets Chinese manufacturers who control 60% of the market, but it also throttles US-headquartered brands like Google Nest and Netgear that rely on Asian assembly lines.

[Samsung Bridges the Apple Walled Garden with AirDrop] · The Verge In an unprecedented win for cross-platform interoperability, Samsung’s new Galaxy S26 lineup now natively supports Apple’s AirDrop. The integration, rolling out globally this week via Samsung’s Quick Share feature, chips away at one of the iOS ecosystem’s stickiest lock-in mechanisms by allowing seamless file transfers between competing devices.

Also Worth Knowing#

  • Nvidia Claims AGI is Here: CEO Jensen Huang boldly declared that artificial general intelligence has been achieved, embracing a polarizing term that most of his cautious peers have recently tried to avoid.
  • Waymo’s Hostile Freeze: A glaring flaw in robotaxi programming was exposed in San Francisco when a Waymo locked its doors and froze in place during a street attack due to nearby pedestrians, trapping a helpless passenger inside for six minutes.
  • Wikipedia Bans LLMs: English Wikipedia has formally banned the use of large language models for writing and rewriting articles in a necessary bid to fight off the “enshittification” of AI-generated inaccuracies.
  • Musk’s Mixed Week: Elon Musk announced “Terafab,” a massive $20 billion AI chip plant in Austin shared by Tesla, SpaceX, and xAI, but also saw his antitrust lawsuit against boycotting X advertisers dismissed with prejudice.