#ai News & Analysis

Quantum-Secure-By-Construction (QSC): A Paradigm Shift For Post-Quantum Agentic Intelligence

Malice in Agentland: Down the Rabbit Hole of Backdoors in the AI Supply Chain

The Pentagon is planning for AI companies to train on classified data, defense official says

The Pentagon is planning to create secure environments for AI companies to train military-specific versions of their models on classified data. AI models like Anthropic's Claude are already being used in classified settings, including for analyzing targets in Iran, but training on classified data would represent a significant expansion of AI use in defense applications.

$12 billion AI startup founder says future tech giants could operate with fewer than 100 employees

A founder of a $12 billion AI startup predicts that future technology giants will be able to operate with teams of fewer than 100 employees due to AI advances. Current AI-enabled startups are already demonstrating the ability to scale to millions of users while maintaining lean organizational structures.

Uber (UBER) Stock Surges 3% Following Massive Nvidia Robotaxi Partnership Across 28 Cities

Uber stock rose 3% following the announcement of a major partnership with Nvidia to deploy robotaxis across 28 cities by 2028. The rollout will begin with Los Angeles and San Francisco in early 2027, marking a significant expansion of autonomous vehicle technology.

Agentic DAG-Orchestrated Planner Framework for Multi-Modal, Multi-Hop Question Answering in Hybrid Data Lakes

Researchers introduce A.DOT Planner, an AI framework that enables multi-hop question answering across hybrid data lakes containing both structured and unstructured data. The system uses directed acyclic graphs to orchestrate complex queries, achieving 14.8% better accuracy and 10.7% better completeness than existing solutions.

ICaRus: Identical Cache Reuse for Efficient Multi Model Inference

ICaRus introduces a novel architecture enabling multiple AI models to share identical Key-Value (KV) caches, addressing memory explosion issues in multi-model inference systems. The solution achieves up to 11.1x lower latency and 3.8x higher throughput by allowing cross-model cache reuse while maintaining comparable accuracy to task-specific fine-tuned models.

Rationale-Enhanced Decoding for Multi-modal Chain-of-Thought

Researchers have developed rationale-enhanced decoding (RED), a new inference-time strategy that improves chain-of-thought reasoning in large vision-language models. The method addresses the problem where LVLMs ignore generated rationales by harmonizing visual and rationale information during decoding, showing consistent improvements across multiple benchmarks.

OrthoFormer: Instrumental Variable Estimation in Transformer Hidden States via Neural Control Functions

Researchers propose OrthoFormer, a new Transformer architecture that addresses causal learning limitations by embedding instrumental variable estimation directly into neural networks. The framework aims to distinguish between spurious correlations and true causal mechanisms, potentially improving AI model robustness and reliability under distribution shifts.

AI is reviving tech sectors that VCs had all but forgotten

According to PitchBook data, AI is driving a resurgence of early-stage venture capital investment into previously neglected tech sectors. Healthcare technology, cybersecurity, biotech, and Software-as-a-Service (SaaS) are experiencing significant funding increases as AI applications revitalize these markets.

Is AI Killing Bitcoin Mining? Here’s The Truth

A debate is emerging over whether AI data centers, which can pay more for electricity than Bitcoin miners, pose a threat to Bitcoin's long-term security. Market and energy specialists are pushing back against claims that this dynamic threatens Bitcoin mining viability.

BlockDAG News: As JPMorgan Gets Dragged Into a $328M Crypto Mess, Traders Dump BDAG & Pepeto and Pour $2.1M Into DeepSnitch AI For 100x

The $32B acquisition that one VC is calling the ‘Deal of the Decade’

Rogue AI agents can work together to hack systems and steal secrets

Forget basketball. Next week’s Nvidia GTC is the real March Madness for AI

3 Questions: On the future of AI and the mathematical and physical sciences

MIT Professor Jesse Thaler outlines a vision for creating a bidirectional relationship between artificial intelligence and mathematical/physical sciences. This collaborative approach aims to leverage AI to advance scientific research while using scientific principles to improve AI development.

An Empirical Study and Theoretical Explanation on Task-Level Model-Merging Collapse

Researchers have identified a phenomenon called 'merging collapse' where combining independently fine-tuned large language models leads to catastrophic performance degradation. The study reveals that representational incompatibility between tasks, rather than parameter conflicts, is the primary cause of merging failures.

AI can rewrite open source code—but can it rewrite the license, too?

The article explores the legal complexities surrounding AI's ability to rewrite open source code and whether such modifications constitute legitimate reverse engineering or create derivative works that must comply with original licensing terms. This raises important questions about intellectual property rights and licensing obligations in AI-generated code.

Legora reaches $5.55 billion valuation as AI legaltech boom endures

Legora, an AI platform for lawyers, has reached a $5.55 billion valuation after raising $550 million in Series D funding led by Accel. The funding round highlights the continued growth and investor confidence in AI-powered legal technology solutions.

Nvidia backs Thinking Machines Lab in multiyear strategic partnership

Nvidia has entered into a multiyear strategic partnership with Thinking Machines Lab, which could accelerate AI advancements and democratize access to cutting-edge AI technology. The partnership is expected to enhance global research collaboration in the AI space.

Anthropic is suing the U.S. government for allegedly blacklisting its AI

Anthropic, the AI company behind Claude, has filed a lawsuit against multiple U.S. federal agencies claiming it was improperly blacklisted from government procurement contracts. The company argues that it was excluded without following the proper legal procedures required to ban a vendor from federal contracting opportunities.

The Download: AI’s role in the Iran war, and an escalating legal fight

This article discusses AI's role in the Iran conflict, specifically how AI models like Claude are being used by the US military for decision-making purposes. The piece appears to be part of a technology newsletter covering AI applications in geopolitical contexts.

How AI is turning the Iran conflict into theater

AI-powered intelligence dashboards are transforming how people consume and experience real-time conflict information, turning serious geopolitical events like the Iran conflict into entertainment-like viewing experiences. The technology enables public access to military intelligence data in ways that gamify and spectacularize warfare.

The ‘Bayesian’ Upgrade: Why Google AI’s New Teaching Method is the Key to LLM Reasoning

Google researchers have developed a new 'Bayesian' teaching method to improve Large Language Models' probabilistic reasoning capabilities. Current LLMs struggle with updating beliefs based on new evidence, falling short in logical reasoning tasks that require maintaining and updating probability assessments.

Agentic retrieval-augmented reasoning reshapes collective reliability under model variability in radiology question answering

Researchers evaluated 34 large language models on radiology questions, finding that agentic retrieval-augmented reasoning systems improve consensus and reliability across different AI models. The study shows these systems reduce decision variability between models and increase robust correctness, though 72% of incorrect outputs still carried moderate to high clinical severity.