Policy

26 Meta employees sued the company, alleging its AI systems targeted workers on medical or family leave for layoffs, violating laws protecting pregnant, disabled, and on-leave employees. Meta denies the claims, saying workforce decisions are made by people, not AI.

The article addresses the challenge of proving ROI for agentic AI in financial services, noting that traditional monitoring fails with multi-agent systems' dynamic costs. Using two real-world use cases—RFP processing automation and AML compliance monitoring—it demonstrates how combining LangChain's observability tools (LangSmith, LangGraph) with Pay-i's economic intelligence platform connects engineering metrics to business value, enabling leadership to see clear returns on AI investments.

Apple is suing OpenAI. The complaint is readable and intense, as these things often are, though many experts seem to think many of the allegations are just the ways things are done. So what does Apple really want here, and why is it picking such a public fight with OpenAI? On this episode of The Vergecast, Nilay and David go through the lawsuit, and look at Apple's history of splashy litigation to determine whether Apple is worried about a possible competitor or simply looking to capitalize on a weak moment for OpenAI. All this is happening as Apple ships the public betas of its new software, headlined by the new Siri AI, and we have thoughts about what it all means — and whether the new Siri is actually any good.

Assistant Professor Bailey Flanigan has arrived at complex computational methods for helping democracy thrive.

Coach’s Corner is a Databricks App that transforms 25 fps match tracking data into a sub-second 2D/3D tactical bench with replays, event analytics, a scout chat, and an opponent-dossier agent. It runs on one platform, powered by Databricks end-to-end: Lakeflow pipelines refine 51 million rows through bronze, silver, and gold; DBSQL queries them in 1-3 seconds; and Lakebase serves them to the app in milliseconds. The AI layer is grounded in governed data, including a Genie space for scouting questions, Vector Search for similar players, and an agentic dossier that calls an LLM served through the Unity AI Gateway, with every step traced in MLflow.

This post provides a high-level overview of the Smartsheet remote MCP architecture, focusing on the AWS infrastructure behind it, including security, governance, scaling, deployment, and AI-specific optimizations.

Startup Factory is an open-source framework that turns project management boards into a governed delivery system for AI agents. It supports multiple trackers, provides layered safety boundaries, and enables deterministic orchestration of cross-functional AI teams.

The article argues that AI memory is the new vendor lock-in, with no real portability existing in July 2026. It identifies three types of lock-in (behavioral, context, relationship), praises early movers like Cognee and ByteRover, but stresses that a neutral interchange standard is needed, as single-vendor formats are just dialects. Regulatory pressure in Europe may accelerate the need.

Bunkerhill Health has raised $55 million to scale its agentic AI platform, Carebricks. The platform is already live at Cleveland Clinic, UTMB, and Intermountain Health. UTMB has deployed over 20 agents across clinical, operational, and administrative workflows, reporting early wins such as a coronary calcium detection agent that flagged a patient at imminent heart attack risk, leading to a life-saving triple bypass.

As President Trump accuses China of stealing US election data, President Xi counters at Shanghai's AI summit, positioning China as a responsible global leader in AI. The event highlights deepening US-China tech rivalry, with China pushing for global AI governance and launching a new international AI cooperation body.

DoorDash releases a command-line interface (dd-cli) enabling AI agents to place real orders on its platform without human approval. While this empowers developers, it sparks debate about disintermediation and DoorDash's business model. Experts warn that refusing to offer such an API could be riskier if agent-led ordering becomes the norm.

Organizations are moving beyond AI deployment to focus on measurable business value, workflow redesign and the governance needed to successfully scale AI.

AI has transformed many industries but made little progress in education because learning requires meaning before mechanics, with a caring human in the loop. The article proposes a two-track education approach: a curriculum track for traditional paths and a child-led track for interests. It emphasizes focusing on meaning, real-world projects, and cognitive apprenticeships enabled by AI.

Experts warn that agentic AI will disrupt enterprise software revenue models, but the 'SaaS apocalypse' is overrated. Providers are focusing on core capabilities to survive disintermediation.

BrowserAct is a CLI tool for AI agents that bypasses anti-bot measures, allows human handoff, runs parallel tasks without interference, and isolates multiple accounts. It features three progressive anti-blocking layers, three browser modes, zero-interference concurrency, and output optimized for LLM reasoning.

This article curates five free resources for learning agentic AI, from structured courses to theoretical foundations and practical evaluation, helping developers build and understand agents effectively.

A recent analysis of AI models from different countries reveals heavy regional censorship on sensitive topics. The author proposes a voluntary international certification standard for AI ethics and transparency to prioritize truth over political interests.

The article argues against both zero-spec and over-specification in agentic development, advocating for a balanced approach with executable checks. It emphasizes that the bottleneck has shifted to defining correctness, and the right amount of specification depends on the task type—exploratory, bounded, deterministic, or multi-agent.

This article summarizes five recent studies on AI in software engineering, revealing that AI compresses upstream work but creates downstream bottlenecks. Key findings: GitHub Copilot increases PR throughput by ~40% with a dose-response effect; AI coding gains (up to +180%) attenuate dramatically through the delivery process (only +30% more releases); productivity and developer experience decouple over time; developers want AI for verification tasks rather than code generation; and cognitive debt and intent debt are emerging as critical software health concerns alongside technical debt.

JetBrains Research explores how AI combined with Extended Reality (XR) can create new interaction paradigms for tech creators. Through expert interviews, they identified five themes: communicating intent to AI-XR systems, AI making XR environments adaptive, barriers to mainstream adoption, changes in creation workflows, and privacy/ethical risks. The study suggests that the convergence of XR hardware and AI may revolutionize technology creation, though technical, cognitive, and organizational constraints remain.

This dataset provides a single-file, pre-embedded SQLite corpus of the EU AI Act (Regulation (EU) 2024/1689), chunked by legal structure with BGE-M3 dense embeddings, metadata, risk tier labels, and more. It is designed for local query and RAG research, with verified completeness and transparent derivation rules.

Einstein’s field equations, Newton’s universal law and artificial intelligence are among the subjects of Laidlow’s ambitious orchestral works on this NMC debut album.

The European Union has issued two new rules requiring Google to share search data and open its Android operating system to rival AI companies, aiming to foster competition and innovation. Google warns the move could undermine user privacy and security.

This paper proposes LIFT, a force-aware post-training framework that adds contact reactivity to pretrained vision-language-action (VLA) policies. By grafting a reactive action expert, injecting 6D end-effector force via causal force memory and cross attention, and coupling with an online DAgger loop, LIFT outperforms vision-only post-training in towel folding, book insertion, and Hanoi ring placement.

This work introduces a multi-modal orchestration framework for semantic audio-driven humanoid control, enabling real-time autonomous selection of motion skills based on music or speech input. Validated on the Unitree G1 humanoid, it demonstrates robust sim-to-real transfer.

Researchers propose an adaptive control method for flexible joint robots with uncertain joint stiffness. The approach updates estimates of nonlinear torque-deflection relations using an implicit control law and a control-input-dependent regressor matrix, and analyzes robustness against motor position controller errors. Experiments on a flexible joint with nonlinear stiffness validate the approach.

SD-MAR is a framework for training and evaluating vision-language models (VLMs) on multi-image analytical reasoning tasks. It constructs paired visual scenarios through controlled perturbations and generates reasoning tasks spanning semantic change attribution and quantitative comparison. Using GRPO-lite with Backward Discounted Allocation (BDA), a reinforcement learning approach that removes KL regularization, fine-tuning on SD-MAR improves in-domain accuracy by up to 36.95% on Qwen2.5-VL-7B and InternVL3-8B. Qwen2.5-VL-7B outperforms GPT-4.1 on the SD-MAR benchmark. Out-of-domain generalization is preserved or improved, with performance within 1% on MME, MMMU-Pro, MathVista and up to 4% improvement on MMBench. LLM-as-judge evaluation shows consistent improvements in logical coherence and explanation quality.

This paper proposes SIRUS, a training-free inference-time framework for concept-level unlearning in text-to-video (T2V) models. SIRUS localizes target-related prompt evidence and suppresses target expression during sampling without updating the text encoder or denoising network. A video-oriented evaluation framework is introduced to separately measure target forgetting, non-target preservation, video quality, jailbreak robustness, and efficiency. On CogVideoX, SIRUS achieves 70.4% average forgetting success and 25.7% average frame hit, compared to 44.4%/47.2% for VideoEraser, while reducing the average VBench quality drop from -0.043 to -0.016. Transfer experiments on Wan2.2 suggest SIRUS generalizes across modern T2V backbones.

The Just Keep Prompting (JKP) framework tests VLM stability under repeated challenging. Evaluations on GPT-4o, Gemini 2.5 Pro, and Qwen3-VL-30B show substantial instability and answer flipping, with model-specific pressure-response profiles.

This paper analyzes why closed-loop knowledge systems (e.g., LLMs, RL) saturate under repeated internal feedback and introduces a three-level operational framework to enable escape via structural interventions. Using Lyapunov drift, stability is characterized, and escape is quantified by attractor displacement and a KL lower bound. Case studies include LLM code repair, sparse-reward RL, and Bayesian optimization.

This paper introduces C3R, a drop-in control layer that, from an inferred domain posterior and no query-time label, certifies a per-domain contamination budget where feasible and otherwise abstains. It guarantees a reduction on the hardest domains, shows stability across resampling, and retains more recall than calibrated cascades.

A new approach called LLM-T1D combines reinforcement learning with large language models to create an interpretable insulin pump controller for Type 1 Diabetes, achieving 73.5% Time in Range while providing clear explanations.

A new paper introduces a three-level hierarchical learning architecture for UAV swarms in search and rescue, integrating Hebbian neuroplasticity, multi-agent RL with GNN and behavior trees, and meta-learning with BDI reasoning. The framework provides formal guarantees and introduces Swarm Meta Cognition.

AI writes code faster than humans can review it, creating a massive trust crisis. Unit tests and prompt engineering aren't enough. Here propose Semantic Contracts—a type-safe, compile-time blueprint that sits between your requirements/prpmpts and code, guaranteeing correctness no matter who (or what) wrote the implementation.

As enterprises race to scale AI, the biggest obstacle to performance and ROI may be the infrastructure moving data, not the hardware processing it. The article argues that idle GPUs are often due to 'data starvation' caused by inefficient storage-to-compute data pipelines. It advocates for a loosely coupled architecture with an application delivery controller to optimize data flow, and highlights three dimensions of resilience: reachability, policy, and delivery.

Gradle Technologies has rebranded to Develocity, evolving its focus to AI-driven software delivery. The company notes that AI has shifted the bottleneck from human developers to the pipeline, requiring new governance and efficiency measures.

Preempt AI v2 is a security standard for AI applications, using ML to defend against prompt injection, jailbreaks, and data leaks. It offers 99.65% accuracy, supports 12+ languages, and has sub-10ms latency.

Meta launched Muse Image, a new AI tool, but faced backlash over a feature that let users tag others to generate AI images using their public photos. The company has disabled the feature, but users must still manually opt out to prevent their photos from being used.

VulnHunter is an open-source, agentic AI security tool that applies proactive, attacker-first analysis directly to source code. It identifies exploitable vulnerabilities, reduces false positives, and provides evidence-backed fixes.

Microsoft's Foundry platform now supports over 80,000 enterprises building AI agents. In an interview, VP Marco Casalaina explains the critical difference between prototypes and production agents, the importance of the agent harness, and how Microsoft builds context layers for reliable agents.

Astrio releases Forall (∀), a coding agent that generates code alongside machine-checkable proofs from user-written specifications. Available as a full CLI agent or a verify-only MCP integration, it currently supports TypeScript, Java, and Rust under the Apache-2.0 license.

The US Copyright Office rules that AI-generated content cannot be copyrighted. An author faces rejection of his book's copyright because he did not preserve the initial AI-written portions, making it impossible to prove human authorship.

Rootly AI Labs introduces Doom Agent Arena, an open-source benchmark using the classic game to test LLMs' reasoning, adaptation, and decision-making skills in dynamic environments. Findings show longer deliberation doesn't guarantee better outcomes, agents can write their own runbooks for efficiency, and speed compounds even if it doesn't win games—offering lessons for AI-assisted incident response.

An in-depth look at the historical Luddite movement—who they were, what they did, and whether they succeeded—and why the modern anti-AI movement cannot simply copy their tactics. The author argues that fundamental differences in context, locality, and specific demands make Luddism a poor model for today's AI resistance.

A new paper introduces MemDecay, a training-free region-aware KV cache eviction policy for LLM agents. It assigns region-specific priorities and decay rates, preserving critical information under fixed cache budget. Experiments show system tokens have much longer half-lives than scratchpad tokens, and pinning system regions retains perfect accuracy where baselines fail.

OpenAI's GPT-Red uses human-AI collaboration for red teaming, a novel approach to model safety, but enterprises must still ensure alignment with their workflows.

The article challenges the narrative that AI distillation by Chinese labs amounts to model theft, arguing that current IP laws do not support such claims. It recommends policy focused on securing access to frontier models rather than expanding IP protections.

Dotmatics Luma and Databricks integrate to harmonize scientific data from instruments, creating a continuous, FAIR-compliant pipeline that enables trustworthy AI in research.

xAI's Grok 4.3 is now generally available on Amazon Bedrock, offering configurable reasoning effort, strong tool use, instruction following, and a 1 million token context window for agentic and enterprise workloads. This post covers its features, access methods, and how to use key capabilities such as chat, reasoning, tool calling, structured output, image input, and multi-turn conversations.

OpenAI trained GPT-Red, an internal-only attacker model, using self-play reinforcement learning against a population of defender LLMs. It beat human red-teamers 84% to 13% on a replicated indirect prompt injection arena, found a novel "Fake Chain-of-Thought" attack class, and cut GPT-5.6 Sol's failures 6x on OpenAI's hardest direct injection benchmark. OpenAI concedes it still struggles with multi-turn and image-based attacks.