Anthropic Launches Bloom: An Open-Source Framework for AI Model Evaluation Anthropic has launched Bloom, an open-source agentic framework engineered to automate behavioral evaluations of advanced AI models. This innovative tool enables researchers to define a specific behavior and then automatically generate numerous test scenarios to measure its frequency and severity. The primary goal of Bloom is to accelerate the creation of high-quality safety and alignment evaluations, which are often labor-intensive and can quickly become obsolete as AI models advance. Anthropic has already utilized Bloom to build and release benchmark results for four key alignment-relevant behaviors across 16 different frontier models. The framework also integrates with popular tools like Weights and Biases for large-scale experiments and complements Anthropic’s other open-source evaluation tool, Petri. ...

Unitree G1 Humanoid Robots Showcase Agility in Onstage Concert Performance In a stunning display of robotic agility, six Unitree G1 humanoid robots from Hangzhou-based Unitree Robotics performed as backup dancers at a Wang Leehom concert in Chengdu, China. Dressed in silver sequin outfits, the advanced robots executed complex, synchronized dance routines and a collective backflip, perfectly timed to the music. This AI-powered performance, praised by Tesla CEO Elon Musk, marks a significant milestone in demonstrating the real-world capabilities and commercial potential of humanoid robotics. Industry observers highlight the event as a major leap forward in robotic dynamic balance and power systems. The G1 robot features 3D LiDAR and depth cameras for advanced perception and boasts a wide range of motion with its 23-43 joint motors. ...

AI Breakthrough: New Molecule Boosts Pancreatic Cancer Chemotherapy In a significant advancement for AI in healthcare, researchers have developed a molecule using artificial intelligence that dramatically improves chemotherapy effectiveness for pancreatic cancer. The AI-generated aptamer, named Apt1, works by disrupting DNA repair mechanisms within cancer cells. This targeted action makes tumors more susceptible to chemotherapy drugs like olaparib, even at lower dosages. This breakthrough highlights the immense potential of AI in drug discovery and creating novel treatments for aggressive cancers by precisely targeting cellular processes. ...

AI and Human Collaboration Boosts Self-Driving Lab Performance by 150% Researchers at Argonne National Laboratory and the University of Chicago have developed a new shared control framework for self-driving laboratories, where human scientists and AI collaborate on the discovery process. This innovative approach has already yielded a 150% increase in mix conducting performance for MIECPs over previous methods. In this model, the AI advisor suggests potential actions, but the human researcher makes the final decision. The team’s future goal is to create a tighter feedback loop where the AI can learn directly from human choices, refining its own decision-making models to better mimic expert intuition. ...

Aptiv & Vecna Robotics Partner on Next-Gen Autonomous Mobile Robots (AMRs) for Warehouses Global industrial technology company Aptiv has announced a strategic collaboration with Vecna Robotics, a leader in AI-driven autonomous material handling. The two companies will co-develop next-generation Autonomous Mobile Robot (AMR) solutions aimed at providing cost-effective automation for warehouses and factories. This partnership will integrate Aptiv’s advanced perception and machine learning technologies into Vecna Robotics’ autonomy and orchestration platform. The goal is to create safer, more efficient, and scalable material handling systems. The collaboration will utilize Aptiv’s PULSE Sensor for 360-degree sensing and its ML-based technologies for real-time perception and path planning. Vecna’s platform will contribute to workflow orchestration and safe navigation in complex environments. Aptiv is set to showcase Vecna Robotics’ CPJ Co-Bot Pallet Jack at CES 2026. ...

Robotics & Autonomous Vehicles Researchers Unveil World’s Smallest Programmable Micro-Robots Researchers at the University of Pennsylvania and the University of Michigan have created the world’s smallest programmable, autonomous robots. These microscopic swimming bots, barely visible to the naked eye, can be programmed with light pulses to move in complex patterns and sense their environment. Each robot, measuring approximately 0.2 by 0.3 millimeters, has its own unique identifier for receiving individual commands. The current models are equipped with sensors that can detect temperature variations, which could be used to monitor the health of individual cells. This breakthrough in micro-robotics opens up new possibilities in fields such as medicine and micro-scale manufacturing. ...

BNP Paribas Deploys Internal AI Tool to Streamline Investment Banking BNP Paribas has introduced a proprietary AI-powered tool, IB Portal, to streamline the creation of client pitches for its investment banking division. The tool operates as a sophisticated search engine, scanning the bank’s historical pitch materials to identify relevant slides and analyses, aiming to reduce repetitive tasks. According to the bank, this innovation can shorten research time by days, enabling teams to focus more on strategic development. The system is built on the bank’s internal “LLM as a Service” platform, which leverages multiple large language models within a secure, private infrastructure. All information generated by the tool requires human review to ensure accuracy and the proper handling of sensitive data. ...

GMO AI & Robotics Acquires Various Robotics to Spearhead Humanoid Robot Development in Japan GMO AI & Robotics Trading Co., Ltd. has acquired all shares of Various Robotics Inc., a developer of advanced robot solutions, as of Monday, December 15, 2025. This strategic move aims to strengthen GMO AIR’s presence in physical AI by uniting some of Japan’s top robotics engineers. The acquisition is set to accelerate the adoption of humanoid robots to address critical social challenges like labor shortages. The companies will jointly develop advanced robot solutions focusing on AI control, manipulation, and autonomous movement. Following the acquisition, Various Robotics will be renamed ‘GMO Various Robotics, Inc.’. The industry has designated 2026 as the ‘Year of Humanoids,’ signaling a significant push to deploy these robots in sectors like factories, logistics, hotels, and hospitals. ...

OpenAI’s GPT-5.2 Release Redefines Professional AI Capabilities OpenAI has released GPT-5.2, its latest and most advanced AI model, specifically engineered for professional knowledge work and enhancing economic value. The new model demonstrates significant improvements in tasks like creating spreadsheets, building presentations, and writing code. GPT-5.2 also showcases enhanced capabilities in image perception, understanding long contexts, and managing complex, multi-step projects. According to OpenAI, the model sets a new standard across numerous benchmarks, outperforming industry professionals in specified knowledge work tasks across 44 different occupations. This release is a strategic response to recent advancements from competitors like Google’s Gemini and Anthropic’s Claude, intensifying the AI landscape competition. The model is being rolled out in tiers, starting with paid plans in ChatGPT, and is available to all developers via the API, marking a significant step in OpenAI’s strategy to accelerate AI adoption in creative, analytical, and operational roles. ...

OpenAI Warns Future AI Models Pose ‘High’ Cybersecurity Risks OpenAI has issued a stark warning that its upcoming frontier AI models could present ‘High’ level cybersecurity risks. The company stated that these advanced models might be capable of automating the discovery and exploitation of significant cyber vulnerabilities, including the development of functional zero-day exploits. To mitigate these potential threats, OpenAI is implementing a ‘defense-in-depth’ strategy. This includes training its models to refuse harmful requests while remaining useful for defensive and educational purposes. The company is also establishing a Frontier Risk Council, an advisory board of cybersecurity experts, to help manage these advanced AI risks. Furthermore, OpenAI is developing a trusted access program to provide enhanced AI capabilities to qualified cybersecurity professionals for defensive operations. ...