Top 10 AI News
US stocks rose again in 2025 after overcoming turbulence from tariffs and Trump’s fight with the Fed
December 23, 2025
This AP report recounts a remarkable year on Wall Street. Despite periodic historic drops over concerns ranging from tariffs to possible AI bubbles, U.S. stock markets ended 2025 higher. Investors faced a “scary” year of volatility but ultimately saw the market recover ([www.investing.com](https://www.investing.com/analysis/a-real-nightmare-before-christmas-266469#:~:text=If%20You%20Thought%20October%20Was,for%20December%E2%80%A6%20Last%20week%20was)).
Google’s parent buys data center energy specialist Intersect for $4.75 billion to help power AI
December 22, 2025
Alphabet announced it will acquire Intersect Power, a company specializing in renewable energy infrastructure, for $4.75 billion. The purchase aims to secure cleaner, more reliable power for the massive data centers that run Google’s AI services. Securing sustainable energy is seen as crucial for supporting the growing demands of artificial intelligence computing.
Starlink in the crosshairs: How Russia could attack Elon Musk’s conquering of space
December 22, 2025
With SpaceX’s Starlink satellites providing global internet, Russia is assessing ways to counter Musk’s space venture. The article explores possible tactics – from jamming signals to physical attacks – that could disrupt the satellite network if Starlink is viewed as aiding Western interests. It highlights the growing risk of geopolitics extending into space technology.
The rise of deepfake cyberbullying poses a growing problem for schools
December 22, 2025
This AP story reports on a disturbing new trend in school cyberbullying. The article describes cases where students used AI tools to generate fake nude images of peers, spreading them online. One middle school girl fought back against those sharing the AI-generated images of her, but was expelled after a schoolyard fight. The incident highlights how deepfake technology is creating unique challenges for education and law enforcement【」】.
Boys at her school shared AI-generated, nude images of her. After a fight, she was the one expelled.
December 22, 2025
An AP video report covers a case in Louisiana where a teenage girl faced harassment from classmates. The girls had used AI to create fake nude photos of her. When she confronted the situation, a fight broke out – and she was expelled, the victim silenced. Authorities had to investigate new school policies on AI-fueled bullying, illustrating how easily generative AI complicates student safety and discipline.
Humanoid robots take center stage at Silicon Valley summit
December 11, 2025
This AP video story highlights the latest in humanoid robot technology showcased in Silicon Valley. Researchers like Hanson Robotics displayed human-like robots that can mimic facial expressions. While these robots generated excitement for their lifelike appearances, experts interviewed remain skeptical about their practical uses. The piece underscores that advanced robotics draws attention, even as questions linger about real-world applications.
Trump signs executive order aiming to block state AI regulations
December 11, 2025
President Trump signed a new executive order that prevents U.S. states from enacting their own artificial intelligence laws. The order asserts federal control over AI regulation, pre-empting state rules in areas like data privacy and algorithm transparency. The move highlights the growing political focus on AI governance, aiming to create a unified national policy rather than a patchwork of state regulations.
MIT in the media: 2025 in review
December 22, 2025
In this end-of-year review, MIT News highlights key stories and research from 2025 that made headlines. The piece covers a wide array of high-impact work by MIT faculty and students, including breakthroughs in artificial intelligence, climate research, and public policy. It offers an overview of how MIT’s innovations reached the public and emphasizes the Institute’s role in addressing global challenges.
Guided learning lets “untrainable” neural networks realize their potential
December 18, 2025
MIT CSAIL researchers developed a novel “guided learning” technique that allows certain neural network architectures – previously thought too unstable to train – to learn effectively. By using a teacher network’s built-in biases to guide them, these “untrainable” networks can now be trained to perform tasks well. This breakthrough could open up new neural designs that were previously abandoned.
A new way to increase the capabilities of large language models
December 17, 2025
Researchers from the MIT-IBM Watson AI Lab introduced an expressive neural architecture that improves large language models’ memory and reasoning over long documents. The new design, reported by MIT News, provides better state tracking across sequences, which helps LLMs answer complex questions about lengthy texts more accurately. This approach could enhance AI understanding in applications like summarization.
A “scientific sandbox” lets researchers explore the evolution of vision systems
December 17, 2025
MIT News reports on a new AI-powered platform that simulates the evolutionary development of visual systems. This “scientific sandbox” allows researchers to model how eyes and vision-related neural circuits evolve under different environmental conditions. The tool could inform the design of better cameras and sensors for robots by mimicking natural vision development.
“Robot, make me a chair”
December 16, 2025
MIT News covers an AI-driven system that lets users create simple objects by describing them in natural language. In the demonstration, participants asked the system to “make a chair” and an autonomous robot assembled a multicomponent lattice chair accordingly. This speech-to-fab prototype combines 3D generative AI with robotics to turn spoken descriptions into physical objects on demand.
3 Questions: Using computation to study the world’s best single-celled chemists
December 15, 2025
MIT News interviews assistant professor Yunha Hwang, whose computational biology research decodes how microbes perform chemistry. Hwang is using AI and genome analysis to understand metabolic networks in single-celled organisms. The article highlights how MIT’s new focus on computation is enabling insights into biology, reflecting the merging of genetics and AI in research.
Deep-learning model predicts how fruit flies form, cell by cell
December 15, 2025
MIT researchers created a deep-learning model that simulates the cellular development of a fruit fly embryo. By predicting how cells arrange and divide during early formation, the AI tool could help scientists detect developmental anomalies. This modeling approach may extend to more complex tissues, helping identify early signals of disease in growing organisms.
Enabling small language models to solve complex reasoning tasks
December 12, 2025
MIT News describes a new “self-steering” system (DisCIPL) that gets smaller language models to work together on tasks with constraints. In demonstrations like itinerary planning and budgeting, multiple AI agents coordinate, passing cues among themselves to handle complexity. This approach shows even lightweight models can tackle difficult problems by collaboratively following defined rules.
New MIT program to train military leaders for the AI age
December 12, 2025
MIT has launched an accredited certificate program to equip U.S. Navy leaders with practical AI skills. Under this new initiative, Naval officers will take courses in AI, data analysis, and technology strategy. The goal is to prepare military decision-makers to integrate advanced AI solutions in operations and defense planning, reflecting how key AI expertise is becoming vital for national security.
New method improves the reliability of statistical estimations
December 12, 2025
MIT News reports that researchers have developed a new statistical technique to quantify the trustworthiness of experimental results. This method assesses when to believe significant findings by improving how scientists interpret p-values and sample data. The advance will help economists, health experts and others decide when study results are reliable or possibly due to random fluctuations.
New materials could boost the energy efficiency of microelectronics
December 11, 2025
This MIT News story covers a breakthrough in chip design that could dramatically reduce power consumption. Engineers stacked multiple active material layers on the chip’s back end, using a new transistor design. By channeling electric current through these layers more efficiently, the approach slashes the energy wasted during computation, marking a step toward greener electronics.
MIT affiliates named 2025 Schmidt Sciences AI2050 Fellows
December 8, 2025
MIT News announced that two MIT-affiliated researchers – postdoc Zongyi Li and Associate Professor Tess Smidt – were named 2025 AI2050 Fellows through Schmidt Futures. Along with other alumni, they will receive funding to pursue ambitious AI research aimed at solving humanity’s toughest problems. Schmidt’s program supports innovators developing AI tools in fields like climate, health, and education.
AI Minted More Than 50 New Billionaires In 2025
December 25, 2025
Forbes reports that the booming AI sector in 2025 created over 50 new billionaires. Investments in AI startups and technology companies surged this year, propelling dozens of founders and executives to billionaire status. The article highlights how unprecedented market growth around AI ventures dramatically increased personal fortunes.
As OpenAI Shifts To For-Profit, Its Foundation Controls $130 Billion. Who Benefits?
December 22, 2025
Contributor Paulo Carvão examines OpenAI’s transition from nonprofit to a for-profit model. With the move, the OpenAI nonprofit board gains control of roughly $130 billion in total assets. The article discusses concerns about transparency and accountability: it asks who will benefit from the foundation’s funds and how this affects AI research and competition.
AI Startup Lovable Just Minted One of Europe’s Youngest Ever Self-Made Billionaires
December 19, 2025
Alicia Park reports on Lovable, a French AI startup that enables users to create branded content. Its founder, just 18 years old, became one of Europe’s youngest self-made billionaires after the company’s latest funding round. The story highlights how rapidly burgeoning AI content platforms are creating huge wealth for their creators.
The Healthcare Industry Outlook For 2026
December 17, 2025
Forbes contributor Amy Feldman outlines key trends for healthcare in 2026. She highlights how AI will continue revolutionizing diagnostics, telemedicine, and drug development. The outlook covers rising telehealth adoption, personalized medicine, and the impact of health tech startups entering the market, predicting significant industry growth and innovation next year.
Let Your Personal AI Shopping Agent Pick The Perfect Holiday Gift
December 17, 2025
This Forbes piece introduces AI-powered shopping assistants that help consumers find gifts during the holiday season. The article explains how smart agents can use personal preferences and natural language to recommend tailored products. By advising users on gift choices, these AI tools aim to simplify shopping and boost sales on e-commerce platforms.
Amazon’s Robotaxi Unit Launches In San Francisco Without Steering Wheels—Or Fees
November 18, 2025
Reporter Alan Ohnsman covers the debut of Zoox’s robotaxi service in San Francisco. The ride-share vehicles are fully driverless—no steering wheels, pedals, or mirrors—and currently free to ride. Zoox co-founder Jesse Levinson anticipates receiving full regulatory approval soon. The rollout marks a milestone in autonomous transit leveraging AI for navigation.
Microsoft Confronts Agent Sprawl With Agent 365
November 19, 2025
In this Forbes article, Janakiram MSV examines Microsoft’s new Agent 365 platform. Designed to manage the proliferation of AI bots, Agent 365 aims to organize and control the use of AI agents across Microsoft’s productivity suite. By cataloging agents and streamlining their roles, Microsoft hopes to prevent “agent sprawl” and deliver a more coherent AI experience within Office applications.
Gemini 3 Approaches The Uber-Software Point: AI As The New UI
November 18, 2025
John Koetsier analyzes Google’s Gemini 3 language model, suggesting it brings AI closer to becoming the new user interface for software. Instead of clicking menus, users may interact with apps via natural language AI agents. The article compares this shift to Uber’s impact on taxis: just as Uber replaced traditional hailing, Gemini and similar models are poised to replace conventional software navigation with voice and text commands.
Windows Is Becoming An Operating System For AI Agents
November 18, 2025
In this analysis, technology columnist Tony Bradley argues that Microsoft is transforming Windows into a platform for AI agents. With features like Copilot built-in, Windows will natively run AI tools for tasks within apps. The operating system’s role is evolving from managing files and applications to orchestrating intelligent assistants, making AI an integral part of the user experience at the OS level.
A Look At The Top LLMs Of 2025
December 29, 2025
Contributor John Werner reviews the largest language models of 2025, comparing their strengths, features, and potential applications. The overview discusses advancements in model sizes and capabilities, and predicts how the next wave of AI agents built on these LLMs could transform industries and daily workflows in the coming year.