Welcome to the October 22, 2025 edition of ACM TechNews, providing timely information for computer professionals three times a week.

HEADLINES AT A GLANCE

• The Day Amazon Broke the Internet
• AI Leaders Push to Pause Superintelligence
• Electronic Implant Restores Sight
• Russians' Surrender to Ukrainian Robot Marks Further Transformation of War
• Chile Embodies AI's No-Win Politics
• Wikipedia Says AI Is Causing a Dangerous Decline in Human Visitors
• Big Tech Sues Texas Over Age-Verification Law
• 15 Million Workers vs. Big Tech's AI Rush
• Microchip Tech Protects Vehicles from Laser Attacks
• Waymo Robotaxis Probed over Safety Near School Buses
• U.K. Tech Industry Grad Hiring Drops 46%
• AI Datacenters Building Their Own Power Plants
• Russia Pushes a State-Controlled ‘Super App’ by Sabotaging Rivals

The Day Amazon Broke the Internet

A faulty software update to Amazon Web Services’ Domain Name System early Monday caused one of the most widespread Internet outages in years, disrupting access to popular platforms and business activity worldwide. The issue traced back to DynamoDB, a core Amazon database that temporarily misrouted traffic crippling operations from package sorting to flight scheduling. The outage highlighted the fragility of cloud infrastructure and its far-reaching economic ripple effects.
[ » Read full article *May Require Paid Registration ]

AI Leaders Push to Pause Superintelligence

A group of AI pioneers, policymakers, and public figures is urging a pause on developing “superintelligent” AI systems until they are proven safe and controllable. The call, led by the nonprofit Future of Life Institute, has gathered over 800 signatures, including those of ACM A. M. Turing Award laureates Yoshua Bengio and Geoffrey Hinton. The group warns that AI is advancing too quickly, with insufficient oversight.
[ » Read full article ]

Electronic Implant Restores Sight

A microchip implant has restored central vision in patients suffering incurable sight loss from dry age-related macular degeneration. In a trial of 38 elderly participants, 84% regained the ability to read letters, numbers, and words with the implant, which is placed under the retina and works in conjunction with augmented reality glasses and a small computer. The glasses project scenes being viewed as an infra-red beam across the chip, which activates the device. AI in the waistband computer processes the information and converts it into an electrical signal, which passes through the cells in the retina and optical nerve into the brain.
[ » Read full article ]

Russians' Surrender to Ukrainian Robot Marks Further Transformation of War

In June, Ukrainian forces in the Kharkiv region used a land drone packed with 138 pounds of explosives to compel Russian soldiers to surrender, the first known capture of prisoners using such a device. The mission showcased how ground-based drones are transforming warfare by taking over dangerous frontline roles from direct assaults to resupply missions. Such remote-controlled robots are becoming vital for Ukraine as it seeks to preserve manpower against Russia’s larger army.
[ » Read full article *May Require Paid Registration ]

Chile Embodies AI's No-Win Politics

Chile is seeking to expand AI capabilities and attract investment, including foreign datacenters, despite scarce resources and public opposition. The South American nation aims to replicate the model implemented in the 1990s to regulate the building and use of foreign astronomical telescopes by providing local universities and companies access to anticipated AI infrastructure in the northern desert city of Antofagasta. Google and Amazon datacenters have faced protests in the capital Santiago over feared environmental harms.
[ » Read full article *May Require Paid Registration ]

Wikipedia Says AI Is Causing a Dangerous Decline in Human Visitors

The Wikimedia Foundation has expressed concern about Wikipedia's long-term sustainability given a steep decrease in human traffic that it attributes to people obtaining information from generative AI chatbots trained on its articles and search engines. The foundation's Marshall Miller said it identified an approximately 8% drop in human traffic in recent months compared to last year after updating its bot detection systems in May in response to unusually high levels of human traffic originating primarily from Brazil.
[ » Read full article ]

Big Tech Sues Texas Over Age-Verification Law

Major tech companies are suing Texas over its new App Store Accountability Act, calling it an unconstitutional “broad censorship regime.” The law, set to take effect Jan. 1, 2026, would require age verification and parental consent for minors to access a wide range of online platforms, from social media and streaming services to educational and news apps. A separate lawsuit filed by student advocates and two minors argues the measure violates the First Amendment.
[ » Read full article ]

15 Million Workers vs. Big Tech's AI Rush

The American Federation of Labor and Congress of Industrial Organizations (AFL-CIO), which represents almost 15 million U.S. workers, has unveiled its "Workers First Initiative on AI," which calls for "worker-centered" AI to protect workers' rights. The union wants human reviews of automated decision-making, transparency in data collection, and a ban on AI as a surveillance tool. It calls for government disclosure of the use of AI in federal systems and more regulation against misinformation campaigns, deepfakes, and other AI abuses.
[ » Read full article ]

Microchip Tech Protects Vehicles from Laser Attacks

Researchers at the French Alternative Energies and Atomic Energy Commission and French semiconductor manufacturing company Soitec developed a new microchip design that uses multi-layer insulation to block laser fault injection (LFI) attacks on vehicles. LFI occurs when a focused laser pulse temporarily disrupts transistors, creating a cyber vulnerability. The Fully Depleted Silicon-on-Insulator (FD-SOI) design isolates transistors with a multi-layer substrate and insulating buried oxide (BOX) layer, making LFI attacks far more difficult.
[ » Read full article ]

Waymo Robotaxis Probed over Safety Near School Buses

The U.S. National Highway Traffic Safety Administration opened a preliminary investigation into about 2,000 Waymo self-driving vehicles after reports that some robotaxis may have failed to obey traffic laws in the vicinity of stopped school buses. The probe follows a media report that a Waymo vehicle equipped with Waymo’s fifth-generation Automated Driving System and operating without a human driver drove around a stopped bus with lights and stop arms extended while students were disembarking.
[ » Read full article ]

U.K. Tech Industry Grad Hiring Drops 46%

Graduate hiring in the U.K. tech industry has decreased 46% over the past year, as AI increasingly replaces entry-level tasks once performed by humans. According to the Institute of Student Employers, companies are using AI for coding, data analysis, and digital work, choosing experienced staff over training new graduates. While overall graduate recruitment dropped 8%, tech was among the sectors hit hardest.
[ » Read full article ]

AI Datacenters Building Their Own Power Plants

Tech companies increasingly are constructing their own power generation facilities to run their datacenters. OpenAI and Oracle are tapping into natural gas as part of their $500-billion Stargate project in Texas, and Equinix's San Jose, CA, datacenter uses fuel cells and solar panels to produce its own power. With a backlog of power-grid projects at a time when high-voltage electric lines are at capacity, datacenters in some areas will have to wait until the 2030s to plug into the grid.
[ » Read full article *May Require Paid Registration ]

Russia Pushes a State-Controlled ‘Super App’ by Sabotaging Rivals

Russia is tightening control over its digital space by promoting MAX, a state-run messaging “super app,” while disrupting competitors WhatsApp and Telegram. The Kremlin has restricted voice and video calls on the two foreign platforms, citing “antifraud” efforts, though critics see it as a push toward a government-monitored Internet. MAX, created by state-controlled VK, now boasts over 45 million users and resembles China’s WeChat, integrating messaging, digital IDs, and government services.
[ » Read full article *May Require Paid Registration ]