Last updated: 13 December 2025

• Adobe Integrates ChatGPT Into Apps

  Adobe is adding ChatGPT-powered generative tools into Photoshop and Acrobat Express, letting users create and edit content via natural language prompts directly within the applications to boost creativity and workflow efficiency. Read More

• Ukraine Building Google-AI System

  Ukraine is developing its own national large language model based on Google’s Gemma framework to support military and civilian AI capabilities, reduce reliance on foreign platforms, and bolster digital sovereignty. Read More

• VA Explores AI for Suicide Prevention

  U.S. lawmakers urged the Department of Veterans Affairs to expand use of AI-based tools and predictive analytics to identify veterans at risk of suicide earlier and support intervention efforts as part of the FY26 VA funding bill. Read More

• Experts Slam AI Research Quality

  Researchers and critics are warning that the AI academic field is flooded with poor-quality papers and output, undermining scientific standards and causing a credibility crisis in AI research publishing. Read More

• BenchSci Extends AI-Driven Research

  BenchSci extended its ASCEND collaboration with Merck to support preclinical disease biology research using AI-based tools, aiming to accelerate drug discovery workflows and enhance biological insight generation. Read More

• Pentagon Launches Generative AI Hub

  The Pentagon is rolling out Google’s Gemini models and a new centralized AI portal to help service members and civilians access approved tools, share use cases, and accelerate responsible AI adoption across the DoD. Read More

• Deloitte: AI Goes Physical With Robotics

  Deloitte’s Tech Trends 2026 identifies Physical AI as a major shift, describing how advances in perception, learning, and onboard compute are enabling robots — including humanoids — to operate more autonomously in real-world environments and move from pilots toward early production use. Read More

• EQT Orders 10,000 Humanoid Robots

  EQT has agreed to buy up to 10,000 humanoid robots from a portfolio company, using the deal to test large-scale deployment and validate commercial demand for AI-powered humanoid systems rather than signaling immediate mass adoption. Read More

• CENTCOM's Attack Drone Task Force

  U.S. Central Command has activated Task Force Scorpion Strike, deploying its first dedicated one-way attack drone squadron equipped with Low-Cost Unmanned Combat Attack System (LUCAS) drones in the Middle East to enhance regional deterrence and operational reach. Read More

• Punto Health Raises €2.3M for AI Care

  UK-based Punto Health secured €2.3M to expand its AI-powered dementia care platform, which provides symptom monitoring, personalized care plans, and caregiver support tools. Read More

• Israel Expands AI, Laser Military Tech

  Israel’s Defense Forces are reorganizing their tech units into a new AI Division and related commands while preparing to field the Iron Beam directed-energy laser air defense system by year-end, aiming to bolster future warfare capabilities and counter evolving threats. Read More

• Palmer Luckey Argues AI Ethics in War

  Anduril co-founder Palmer Luckey says using advanced AI in life-and-death military decisions is ethically necessary, arguing that inferior technology cannot provide a moral high ground and emphasizing AI’s role in reducing collateral damage. Read More

• Nvidia Can Sell H200 to China

  The U.S. has cleared Nvidia to sell its H200 AI chips to China under export controls, but analysts question whether Chinese buyers will adopt them given restrictions, rising domestic alternatives, and uncertainty around long-term access. Read More

• DeepSeek Unveils V3.2 & Speciale

  Chinese AI developer DeepSeek released V3.2 and V3.2-Speciale, two advanced open models boosting reasoning depth, agent performance, and cost efficiency — with V3.2 delivering “GPT-5 level” daily use performance and the Speciale variant rivaling Google’s top models on benchmarks. Read More

• Z.ai Open-Sources GLM-4.6V Model

  Z.ai released GLM-4.6V, an open-source vision-language model with native tool calling that lets developers build multimodal agents able to reason over images and directly invoke external tools for tasks and workflows. Read More

• Mistral Launches Devstral 2, Vibe CLI

  Mistral released Devstral 2, a coding-focused model, alongside Vibe, a CLI tool that lets developers interact with models directly from the terminal. The update targets faster iteration, tighter developer workflows, and improved code understanding and generation for local and cloud environments. Read More

• IBM Granite Leads Transparency Rankings

  IBM’s Granite family of foundation models ranked #1 in the latest Stanford Foundation Model Transparency Index with the highest score in the index’s history, scoring top marks on domains like data acquisition and governance, highlighting transparency as a differentiator in the AI ecosystem. Read More

• Federal Order Preempts State AI Laws

  The White House issued an EO asserting federal primacy over AI policy, directing agencies to identify and challenge state AI laws deemed obstructive to national objectives. The order aims to prevent a fragmented regulatory landscape and accelerate AI development under a unified federal framework. Read More

• Navy, Palantir Sign $448M AI Deal

  The U.S. Navy awarded Palantir a $448M contract to integrate AI for submarine and shipbuilding programs, using predictive analytics and digital engineering tools to speed production and improve maintenance planning. Read More

• Robot Learns Skills From One Demo

  Texas A&M researchers unveiled a robotic system that can learn new tasks from a single demonstration using fast-adapting AI models, improving robotics’ ability to operate in unstructured environments with minimal training. Read More

• Humanoids Poised for Construction Work

  Construction firms are beginning to test AI-powered humanoid robots for tasks like material handling and repetitive labor, with advances in perception, mobility, and dexterity positioning them as a potential solution to chronic workforce shortages. Read More

• Fei-Fei Li: Spatial Intelligence Frontier

  AI pioneer Fei-Fei Li argues that while current AI excels at text and visual tasks, the next major leap is spatial intelligence — giving machines a deeper understanding of 3D environments — which could transform robotics, virtual worlds, creativity, and scientific discovery. Read More

• OpenAI Releases GPT-5.2 Model

  OpenAI introduced GPT-5.2, an upgrade focused on stronger reasoning reliability, improved multimodal understanding, and better instruction following, particularly for complex, multi-step tasks. The release emphasizes stability, safety tuning, and enterprise readiness rather than a major architectural leap. Read More