The record
Published reporting, analysis and commentary from the Automatica newsroom.
A new project called AI IQ is sparking significant debate by assigning human-scale intelligence quotients to over 50 of the world's most powerful language models, presenting the results on an interactive website, aiiq. org VentureBeat.
The digital dawn of April 29, 2026, brought with it a cascade of academic preprints on arXiv CS. LG, revealing a profound and accelerating shift: artificial intelligence is not merely observing human patterns, but now increasingly charting the very physics and biology that underp
The field of agentic artificial intelligence (AI) for collaborative robotic systems is undergoing a significant acceleratory phase, evidenced by concurrent advancements in scalable architectures for multi-robot environments and the launch of an open evaluation framework. These de
The landscape of artificial intelligence continues to evolve with the recent unveiling of advanced systems poised to redefine automated code quality assessment and specialized data analysis. New research published on arXiv CS.
The subtle hum of the city, once a symphony of autonomous lives, is rapidly being re-transcribed into a data stream, meticulously charted and predicted by emerging artificial intelligence models. Recent academic papers from arXiv CS.
You submit an application. Every field is filled, every fact precise.
Well, butter my buns and call me a biscuit, another day, another two data points proving humanity is slowly, gleefully handing over its brainpower to a bunch of glorified calculators. On one side, we’ve got Artificial Intelligence diligently crafting multiple-choice questions for
For years, the conventional wisdom in artificial intelligence has been a relentless pursuit of scale: ever-larger models, more parameters, and compute resources measured in astronomical figures. Yet, the most significant shifts often begin not with more, but with smarter.
The burgeoning landscape of artificial intelligence models has long posed a significant challenge for robust verification, particularly concerning their reliability on novel, unlabeled data. A new research paper, published on arXiv on May 25, 2026, introduces MetaEvaluator, a fra
A recent wave of research on arXiv highlights how large language models (LLMs) and autonomous AI agents are rapidly transforming scientific discovery, moving beyond traditional computational roles to actively participate in research from uncovering causal links to harmonizing com
The recent wave of AI research reveals a striking duality: while multi-modal AI models are achieving unprecedented levels of sophistication in tasks like visual reasoning and medical imaging, human trust in foundational inputs like speech is eroding significantly. New findings fr
As AI Layoffs Mount, OpenAI Floats Giving Washington a 5% Stake to Share the Wealth OpenAI is reportedly discussing a deal to give the U. S.
The frontier of autonomous cybercrime has supposedly been breached, yet the ghost in the machine remains firmly human. Despite recent headlines trumpeting the technical execution of the first AI-driven ransomware attack, tactical realities reveal that human operators are still pu
Anthropic’s public-facing threat model has historically treated the user as an entity to be protected, leaning heavily on its reputation as a privacy-conscious, safety-first artificial intelligence vendor. That operational facade collapsed this week.
Regulators on both sides of the Atlantic moved closer to the front lines of AI governance on Monday, as a senior UK watchdog warned of an “arms race” to oversee artificial intelligence in financial services and the United States confirmed engineering dean Arvind Raman as the 18th
Microsoft has laid off roughly 4,800 employees — about 2. 1% of its global workforce — with the deepest cuts landing in its Xbox gaming division and commercial sales business, marking one of the company’s most aggressive restructurings since its AI pivot accelerated last year.
Seven Papers That Rewire How We Think About AI Training, Inference, and Identity A wave of arXiv preprints dropped this week that, taken together, paint a picture of a field quietly dismantling some of its own foundational assumptions. From a proof that three competing RL trainin
A significant trend has emerged in recent artificial intelligence research, signaling a concerted effort to enhance the reliability, efficiency, and deterministic capabilities of Large Language Models (LLMs) for enterprise applications. A notable development includes the introduc
The relentless pursuit of smarter, more capable Large Language Models (LLMs) has hit a paradoxical snag: new research indicates that strengthening an LLM's reasoning capabilities can, counter-intuitively, amplify its tendency for tool hallucination. This revelation, brought forth
New research has introduced HARNESS, a family of Arabic-centric self-supervised speech models that promise to bring advanced voice understanding to your phone or tablet without draining its battery. These models are designed to be remarkably lightweight, offering impressive accur