Daily Tech News Digest — Saturday, March 7, 2026

40 articles analyzed · 8 sources · 5 key highlights

Key Highlights

Overview

Saturday's tech news was dominated by corporate confrontations with government, AI ethics controversies, and hardware delays affecting major players. Nintendo filed a lawsuit against the U.S. government seeking tariff refunds, while the administration admitted it lacks the technical infrastructure to process billions in refunds. Meanwhile, AI company Anthropic saw user growth surge following its Pentagon deal collapse, Grammarly faced backlash for using people's identities without permission, and Valve disappointed gamers by suggesting its Steam Machine might not ship in 2026 after all.

Nintendo vs. U.S. Government Over Tariffs

Nintendo of America filed suit in the U.S. Court of International Trade against the Department of Treasury, Department of Homeland Security, and U.S. Customs and Border Protection, seeking refunds on tariffs the Supreme Court recently struck down. The lawsuit cites a February Supreme Court ruling that invalidated some of President Trump's sweeping tariff policies, which had impacted Nintendo and thousands of other companies. Nintendo had already raised Switch prices in August 2025 citing "market conditions" but has kept the newer Switch 2 console price unchanged so far. In a remarkable admission, U.S. Customs and Border Protection revealed it currently can't process the billions of dollars in refunds owed to companies. In a Friday filing, CBP executive director Brandon Lord stated the agency's digital import processing system is "not well suited to a task of this scale." The technical limitation adds another layer of complexity to an already contentious trade situation, leaving companies like Nintendo in limbo despite their legal victories.

Anthropic's Pentagon Fallout Becomes User Windfall

The public breakdown between Anthropic and the Department of Defense over AI ethics has unexpectedly boosted the company's consumer business. After the Pentagon officially designated Anthropic a supply-chain risk when negotiations failed over military control of its AI models—including use in autonomous weapons and mass domestic surveillance—Claude's consumer app is now seeing more new installs than ChatGPT and growing its daily active users. The controversy proved instructive for the startup world: OpenAI accepted the Pentagon's terms that Anthropic rejected, then watched ChatGPT uninstalls surge 295%. Microsoft, Google, and Amazon clarified that Anthropic's Claude remains available to non-defense customers through their platforms, attempting to contain the damage. The episode raises fundamental questions about Pentagon surveillance authority, with legal experts noting that more than a decade after Edward Snowden's revelations, whether the law actually allows the U.S. government to conduct mass surveillance on Americans remains surprisingly unclear.

Grammarly's Identity Problem

Grammarly sparked outrage after journalists discovered the company's "expert review" feature uses people's identities without permission. The Verge's staff found that the AI-generated feedback system offered writing advice "inspired by" their own editors, including editor-in-chief Nilay Patel and other senior staff members. The feature also included recently deceased professors, as Wired reported Wednesday. The controversy highlights growing concerns about AI companies appropriating individuals' professional reputations and expertise to train and market their products without consent or compensation.

Valve's Steam Machine Delay

Valve walked back its hardware launch timeline, now saying it only "hopes" to ship the Steam Machine, Steam Controller, and Steam Frame in 2026—a significant downgrade from previous commitments. In a Year in Review blog post, the company cited ongoing memory and storage chip shortages. As recently as last month, Valve had explicitly stated it hadn't changed plans to ship all three products "in the first half of 2026." The new hedging language suggests the company is preparing customers for potential delays into 2027, a disappointment for PC gamers anticipating the new hardware.

DJI Pays Security Researcher $30K

DJI agreed to pay $30,000 to Sammy Azdoufal, the man who accidentally discovered he could access 7,000 DJI Romo robot vacuums and peek into other people's homes while simply trying to connect a PlayStation gamepad to his own device. The Valentine's Day discovery made headlines worldwide. While DJI had begun addressing some vulnerabilities before Azdoufal's disclosure to The Verge, the incident exposed serious security flaws in the company's IoT infrastructure and highlighted the risks of connected home devices.

Other Notable Developments

**LLM Code Quality Debate**: Discussion continued around large language models' coding capabilities, with analysis suggesting LLMs write "plausible code" rather than correct code—a distinction with significant implications for software development workflows. **TerraPower Nuclear Approval**: Bill Gates' TerraPower received approval from the Nuclear Regulatory Commission to build a new nuclear reactor, the first such permit issued in nearly a decade, signaling potential momentum for next-generation nuclear energy. **Apple App Store Changes**: Apple users in the U.S. can no longer download ByteDance's Chinese apps following TikTok's transfer of U.S. operations, marking another chapter in the ongoing geopolitical tech decoupling.

Outlook

The convergence of government technical incompetence (CBP's inability to process refunds) and aggressive policy positions (Pentagon surveillance demands) is creating unusual market dynamics where ethical stances can become competitive advantages. As more companies navigate the tension between government contracts and consumer trust, Anthropic's experience may serve as a template—or warning—for the AI industry. Meanwhile, hardware supply chain issues continue plaguing major launches, suggesting the chip shortage's effects will extend well into 2026.