Top 10 AI News
“`
Amazon’s Robotaxi Launches in San Francisco Without Steering Wheels or Fees
By Alan Ohnsman, Forbes Staff – November 18, 2025
Amazon subsidiary Zoox launched a robotaxi service in San Francisco using electric vehicles that lack steering wheels, mirrors, or pedals. The company co-founder Jesse Levinson says rides are free for now, and regulatory approval is expected soon. The development comes as Zoox tests fully autonomous technology on city streets.
Microsoft Launches Agent 365 to Manage AI Agent Sprawl
By Janakiram MSV, Senior Contributor – November 19, 2025
Microsoft introduced “Agent 365”, a new platform designed to centralize and control the proliferation of AI agents in enterprises. The tool aims to help organizations manage security and governance as employees create many custom AI assistants. By offering a unified dashboard, Agent 365 seeks to streamline oversight of AI-driven tools and ensure compliance across the organization.
Gemini 3 Pushes AI as the New User Interface
By John Koetsier, Senior Contributor – November 18, 2025
Google’s Gemini 3 release emphasizes using AI as a natural interface across applications, aiming to transform how users interact with software. The update includes features where the AI model can understand context and visual elements, effectively making AI “the new UI.” Beta users reported being able to use Gemini to automate tasks by simply conversing in natural language with their apps.
Windows Becomes an Operating System for AI Agents
By Tony Bradley, Senior Contributor – November 18, 2025
Microsoft is reimagining Windows to better support AI-driven workflows, effectively turning the OS into a platform for AI agents. New development features in Windows will allow third-party AI bots to run in the background, sharing data and performing tasks for the user. The company envisions a future where Windows seamlessly coordinates multiple AI services to assist with everyday computing needs.
Why America’s Power Grid Will Withstand the $2.5 Trillion AI Datacenter Boom
By Christopher Helman, Forbes Staff – November 27, 2025
A report indicates that the U.S. power grid has sufficient capacity to meet the enormous energy needs of a projected $2.5 trillion investment in AI data centers. Experts argue that a mix of solar, battery storage, natural gas, and next-gen nuclear power can be leveraged. Essentially, they conclude that with strategic planning, the surge in AI-driven demand can be accommodated without major grid failures.
$150 Million AI Lobbying War Fuels Fight Over Preemption
By Paulo Carvão, Contributor – November 28, 2025
Huge sums are flowing into Washington as tech coalitions vie to influence upcoming U.S. AI regulations. With allies of major AI companies spending around $150 million, legislators are embroiled in a battle over federal preemption of state AI laws. Some industry groups want a unified federal framework to avoid a patchwork of state rules, sparking debate over national vs. local regulatory authority.
“Software-as-a-Service” or “Services-as-Software”? A New AI Era
By John Werner, Contributor – November 28, 2025
Industry experts explain a shift from traditional SaaS to a model they call “Services as Software,” where AI automates entire services. Instead of human-delivered services, businesses can use AI-driven platforms to fulfill customer needs end-to-end. The approach envisions a future in which what we know as services (like legal advice or consulting) become software-automated solutions.
Tech Roundup: Snap Hits 1B Users; Meta’s XR Apps; OpenAI and Apple News
By Charlie Fink, Contributor – November 28, 2025
A batch of tech headlines: Snap announced it has nearly 1 billion users, underscoring the ongoing social media surge. Meta unveiled new mixed-reality apps leveraging AI generation for immersive experiences. Apple’s design guru Jony Ive hinted at new AI-related hardware developments. Music startup Suno announced deals with Warner for AI-generated tracks, and Apple added interactive sports content to Vision Pro, highlighting AI-driven innovation across industries.
“Our GPUs Are Melting,” Warns Sora CEO as New Limits Are Imposed
By Barry Collins, Senior Contributor – November 28, 2025
The CEO of Sora, which makes specialized AI chips, cautions that demand is pushing hardware to its limits as if “our GPUs are melting.” The surge in generative AI use has led companies like OpenAI to cap free user access to computationally heavy features. The warning underscores the massive server power required for AI workloads and the industry’s current constraints in meeting that demand.
Meet Casio’s New AI-Powered Pet, Moflin
By Charlie Fink, Contributor – November 28, 2025
Casio introduced Moflin, a robotic pet toy with built-in AI that learns over time. Moflin responds to touch and sound, gradually developing a unique personality through interactions. Unlike some other robot companions, Moflin has a plush, lifelike body but relies on simple lights for expressions; however, its developers say the AI-driven behavior gives it a charming, evolving character.
Study Finds a Pattern Bias That Can Trip Up Large Language Models
MIT News – November 26, 2025
MIT researchers have identified a key issue that can make large language models unreliable. They found that LLMs tend to form accidental associations between specific sentence patterns and topics, so the model might repeat those patterns instead of truly reasoning. This “shortcut” behavior suggests that current AI may be less robust than expected when facing unfamiliar phrasing.
MIT Scientists Develop AI Model to Design Treatment Molecules
MIT News – November 25, 2025
MIT researchers unveiled BoltzGen, a generative AI model that can create new protein-binding molecules from scratch. The system designs potential drug candidates for hard-to-treat diseases by simulating how molecules would bind to biological targets. This work extends AI’s role beyond data analysis into engineering novel compounds for medicine.
How Artificial Intelligence Can Help Achieve a Clean Energy Future
MIT News – November 24, 2025
AI is playing a growing role in advancing renewable energy and sustainability. The article describes how machine learning models are being used to manage power grid operations, optimize infrastructure planning, and design new materials for batteries and solar cells. By making forecasting and system controls smarter, AI tools could help accelerate the transition to cleaner energy systems.
Study Reveals Similarities in Human and AI Problem Solving
MIT News – November 19, 2025
MIT neuroscientists found parallels between how human brains and advanced AI models approach complex tasks. By analyzing performance on certain cognitive problems, they observed that both humans and AI tend to favor straightforward solution paths under time pressure. The work suggests that insights into human cognition may help us better understand AI decision-making patterns.
New AI Agent Learns to Use CAD Software from Simple Sketches
MIT News – November 19, 2025
MIT researchers have developed an AI called VideoCAD that can operate computer-aided design (CAD) tools to create 3D models based on rough user sketches. This AI agent learns from video demonstrations and helps users by automating parts of the design process in software like CAD. The system aims to boost designer productivity and make engineering more accessible.
Understanding the Nuances of Human-Like Intelligence
MIT News – November 11, 2025
Associate Professor Phillip Isola of MIT is investigating how AI “thinks” compared to humans to safely integrate intelligent systems. In an interview, he explains that by studying human perception and reasoning in detail, we can identify where AI might behave unexpectedly. His goal is to develop AI that aligns better with human values and reasoning processes in real-world interaction.
Charting the Future of AI: From Safer Answers to Faster Thinking
MIT News – November 6, 2025
Students involved with the MIT-IBM Watson AI Lab shared how they’re pushing AI in new directions. Projects highlighted include methods to make AI answers more fact-based and less prone to error, as well as techniques to accelerate language model reasoning. The work reflects efforts to improve AI reliability and efficiency for future applications.
Researchers Propose New Model for Legible, Modular Software
MIT News – November 6, 2025
MIT computer scientists introduced a coding framework that makes software easier for AI to understand and generate. By using clear modules and simple synchronization rules, the new model helps AI programmers (like GPT-based code generators) produce code that is more reliable and maintainable. The approach aims to reduce software bugs and improve collaboration between human developers and AI tools.
Teaching Robots to Map Large Environments
MIT News – November 5, 2025
MIT engineers developed a new technique to help robots rapidly build maps of unknown areas, which is crucial for search-and-rescue missions. Using a “grounded transformer” model, a robot can integrate sequential sensor data to predict and construct an accurate map of its surroundings on the fly. This allows a drone or robot to navigate unpredictable environments without human guidance.
The Brain Power Behind Sustainable AI
MIT News – October 24, 2025
MIT graduate student Miranda Schwacke is exploring how the brain’s energy-efficient computing can inspire greener AI. By studying neural circuits, she hopes to design AI hardware and algorithms that consume less power. The goal is to use principles of human brain architecture to drive down the energy footprint of large-scale AI systems.
AI-Generated Song Sparks Ethical Questions in Nashville
AP News – November 28, 2025
Grammy-nominated country artist Blanco Brown says he’s not surprised that an AI-generated song mimicking his style went viral. The track, “Walk My Walk,” was attributed to a fictional artist, and Brown notes that people have been copying his sound since he moved to Nashville. The incident raises new questions about music, attribution, and how artists protect their creative works in the era of AI.
Judge Raises Concerns Over ICE Agents Using AI Translation
AP News – November 26, 2025
A federal judge noted that immigration enforcement agents had used an AI tool to translate evidence from Spanish, and he questioned the accuracy and potential privacy issues. The judge’s footnote highlights concerns about relying on AI for sensitive law enforcement tasks without clear oversight. Legal experts say the case underscores the need for scrutiny of AI in judicial matters.
Consultant Defies Order in Lawsuit Over AI Robocall
AP News – November 25, 2025
A Republican consultant is challenging a court order after an AI-generated robocall impersonating President Biden aired during a New Hampshire campaign event. The call, which used AI to mimic the president’s voice, led to allegations of illegal electioneering. The consultant is defying a summons, setting up a legal showdown over whether AI can be used in political communications.
Alibaba’s Cloud Revenue Soars on AI Demand
AP News – November 25, 2025
Chinese tech giant Alibaba reported a 34% year-over-year jump in cloud services revenue, largely driven by artificial intelligence demand. The growth was fueled by businesses using Alibaba’s cloud AI offerings for large-scale machine learning projects. Investors noted that AI is becoming a key earnings driver for major cloud providers, reflecting strong enterprise interest in the technology.
Trump Signs Executive Order for AI ‘Genesis Mission’
AP News – November 24, 2025
The White House announced the “Genesis Mission,” an AI-focused project under an executive order signed by President Trump. The initiative pools resources across federal agencies to supercharge scientific research with AI, such as drug discovery and climate modeling. The goal is to maintain U.S. leadership in AI-driven innovation by coordinating efforts and funding at a national level.
OpenAI and Taiwan’s Foxconn Partner on AI Hardware
AP News – November 21, 2025
OpenAI has teamed up with Taiwan’s Foxconn to design and manufacture specialized AI chips for the U.S. market. The agreement entails developing new processor designs at OpenAI’s labs and then producing them in Foxconn’s U.S. fabs. The collaboration is aimed at building the next generation of AI hardware to meet growing demands for compute power.
Nvidia Beats Earnings, Sending Global Stocks Higher
AP News – November 20, 2025
Nvidia’s better-than-expected quarterly results sparked a surge in world stock markets, as investors regained faith in an AI-driven growth narrative. The chipmaker reported strong revenue from its data center business, powering optimism that demand for AI hardware remains robust. Analysts said removing doubts about an AI “bubble” helped renew enthusiasm in the tech sector.
Industry Titans Sign AI Music Licensing Deals
AP News – November 20, 2025
Major record labels Sony, Warner, and Universal have struck deals with AI music platform Klay to license their entire music catalogs for AI training. The agreements mark the first time big labels formally allow AI companies to use copyrighted songs to train generative music models. This move aims to address legal uncertainties over AI-generated music and ensure that artists and labels are compensated.
One Tech Tip: Do’s and Don’ts of Using AI for Schoolwork
AP News – November 20, 2025
An educator’s advice column outlines how students can use AI tools like ChatGPT responsibly in their studies. It suggests using AI to get help starting essays or understanding concepts, but cautions against copying answers or submitting AI-generated text without attribution. The piece emphasizes that students should learn from AI assistance rather than rely on it to complete assignments.
Advocacy Groups Urge Parents to Avoid AI Toys This Holiday
AP News – November 20, 2025
Consumer rights groups are warning parents to be cautious about AI-powered toys like Miko and Gabbo for children. They argue that some “smart” toys record voices and conversations, posing privacy and safety risks. The outreach suggests families opt for non-AI gifts this year until these devices have stronger protections against data misuse.
Putin Inspects Robot and Talks AI at Nuclear Plant
AP News – November 19, 2025
Russian President Vladimir Putin toured a manufacturing plant with a robot that greeted him by name, underscoring Russia’s push into robotics. He also discussed plans to incorporate AI (particularly for monitoring and control) into building new nuclear reactors. The visit highlights how Russia is blending advanced technology with its strategic energy projects.
Harvard Opens New Investigation Into Ex-President Summers
AP News – November 19, 2025
Following the release of Jeffrey Epstein-related emails, Harvard University launched a fresh inquiry into Larry Summers, who was president of MIT and later a Harvard professor. The correspondence, which included messages involving Epstein, prompted questions about Summers’ actions and whether university ethics rules were violated. Summers had already agreed to step back from teaching duties amid the controversy.
Nvidia Earnings Cow Fears of an AI Bubble, But Investors Are Cautious
AP News – November 20, 2025
Tech investors are uneasy as Nvidia’s latest financial results reinvigorate the market’s faith in an AI-driven boom but also raise concerns about lofty valuations. While strong sales eased immediate fear of a bubble bursting, analysts note that demand must keep pace to justify sky-high stock prices. The article suggests that, for now, excitement about AI’s potential is being tempered by caution over future performance.
Wall Street Ends Volatile Day Ahead of Key Economic Data
AP News – November 19, 2025
The stock market swung between gains and losses amid jitters about high valuations for AI companies. Nvidia’s stellar earnings earlier in the week had initially boosted confidence, but by Tuesday concerns about inflation and interest rates returned. The Dow, S&P 500, and Nasdaq all held near break-even after the day’s trading, as investors awaited further clues on economic policy.
Mexico Plans Latin America’s Most Powerful Supercomputer
AP News – November 20, 2025
The Mexican government announced plans to build a supercomputer dubbed “Coatlicue” that will be the region’s fastest, with an anticipated processing power rivaling top global machines. Officials say Coatlicue will be named after an Aztec goddess and used for scientific research in fields like climate modeling, health, and AI development. The project aims to boost Latin America’s capacity for advanced computing and innovation.
“`