Artificial intelligence

117 articles

Reporting on AI models, tools, research, infrastructure and policy.

Coverage: AI

The office building at 1515 Third Street in San Francisco's Mission Bay neighborhood, which was OpenAI's headquarters when the photograph was taken.

ANALYSIS AI Policy and Safety

OpenAI proposes the rules for checking its own safety claims: what its third-party assessment framework promises, and what it leaves out

OpenAI's September 22, 2026 framework proposes four priority areas and five principles for independent scrutiny of its safety claims. The substance is detailed and at times candid, but as a voluntary, self-authored document it leaves the hardest questions about independence unanswered.

NVIDIA GeForce RTX 50 Series graphics cards shown in a video frame about the RTX 50 generation

NEWS Gaming and Graphics

NVIDIA's DLSS 5 puts a generative neural model inside the game frame, and NBA 2K27 is the first to ship it

NVIDIA's developer blog says DLSS 5's 3D-Guided Neural Rendering treats the game engine's frame as a fixed foundation and adds lighting and material detail with a deterministic, one-frame-in, one-frame-out model. It is shipping today in NBA 2K27 on GeForce RTX 50 Series GPUs. Here is what the technology does, what NVIDIA claims, and what remains unverified.

Delegates seated at tables with national nameplates during a multilateral session at the World Trade Organization Public Forum in Geneva

NEWS AI Policy

OpenAI Calls for Global Frontier-AI Standards, While Ruling Out Licenses and Mandatory Pre-Release Review

OpenAI's Global Affairs team published "Building standards for the next phase of AI" on September 21, 2026, urging international technical standards for frontier AI and recursive self-improvement through CAISI and the AI safety institutes network, while stating the standards would not be licenses, mandatory prerelease review, or approval requirements.

A person using a smartphone outdoors, illustrating everyday connected technology use.

NEWS AI Policy

OpenAI Publishes Six-Pillar Youth Safety Blueprint for Australia as Teen-Focused ChatGPT Rolls Out

OpenAI's Australian Youth Safety Blueprint, published September 18, 2026, sets out six pillars for protecting young AI users as a default ChatGPT experience for identified teenagers aged 13 to 17 rolls out in Australia. The plan is company-reported and not yet independently evaluated.

Josh Hawley poses in a dark suit and red tie between the United States and Missouri flags.

NEWS Artificial Intelligence

Hawley Opens OpenAI Inquiry Seeking Records on the Hugging Face Agent Breach

The September 10 investigation asks what leaders knew and how testing resumed, while public reports confirm the breach without settling internal responsibility.

A finite-state-machine diagram with five circular states labeled q1 through q5, directed arrows and transitions marked a or b.

ANALYSIS Technology

AWS Lambda Makes Pydantic AI Agent Runs Resumable, Not Exactly Once

The integration checkpoints model and tool calls, reducing repeated work after interruptions while leaving external side effects subject to retries.

Chris Van Hollen poses in a dark suit before the softly focused United States Capitol.

CLAIM CHECK Artificial Intelligence

Van Hollen Calls OpenAI Models the Most Efficient Hackers Ever. Published Tests Do Not Prove It

Astra shows major cyber gains in controlled evaluations, but no public comparison ranks it against every human or automated hacking system.

Official portrait of House Speaker Mike Johnson standing in a dark suit.

CLAIM CHECK Artificial Intelligence

Mike Johnson’s Warning That Rushed AI Rules Would Cost the China Race Is Unproven

Johnson’s forecast concerns an undefined emergency regulatory response, and the CNN interview does not establish that it would cause a U.S. defeat.

A hidden Markov model diagram with three hidden states labeled A, B and C, transition probabilities, and arrows leading to two observed nodes labeled X and Y.

ANALYSIS ARTIFICIAL INTELLIGENCE

Belief-State Engine Separates Uncertainty Tracking From LLM Decisions

An external Bayesian filter gives an LLM a compact posterior instead of a growing history, but the first experiments are small and incomplete.

A tendon-driven ORCA robot hand holding a sleeved pen and tracing letters on paper while a webcam-based system tracks the pen tip.

ANALYSIS Robotics

Robot Hand Starts Writing After an 18-Second Physical Calibration

An ETH Zurich controller learned how finger commands moved a grasped pen without simulation training or demonstrations, achieving a reported 0.64-millimeter mean in-plane error in laboratory tests.

A vertical software-development life-cycle diagram linking planning, requirements analysis, design, implementation, testing, deployment and maintenance with blue arrows.

NEWS Technology

Palantir Adds Isolated Tests for LLM Pipeline Changes

Pipeline Builder can now test Use LLM nodes against selected cases and store row-level results before teams change a production pipeline.

Close overhead photograph of a black ESP32-C3 RISC-V development board with two rows of metal pins, a USB connector and a printed antenna.

ANALYSIS ARTIFICIAL INTELLIGENCE

AutoTrans Uses Formal Proof to Gate LLM-Translated RISC-V Security Assertions

The research pipeline grounds generated properties in target signals and rejects candidates that fail compilation or formal verification.

The illuminated arched entrance and stone facade of 20 Exchange Place in Manhattan.

ANALYSIS Artificial Intelligence

ChatGPT for Financial Services Brings Licensed Data Into the Analysis Workflow

OpenAI’s finance product combines GPT-6 Astra, citations and institutional controls, but integrated data does not guarantee correct analysis or measurable business gains.

A labeled continuous-integration diagram connecting developers, a code repository, build tooling, a CI server and status notifications.

ANALYSIS ARTIFICIAL INTELLIGENCE

GitHub Gives Copilot Code Review Tools to Test Its Findings

Copilot can now run builds and tests and combine several Lite review agents, but GitHub's measurements do not establish defect-detection accuracy.

A stylized blue human head and brain appear over a dense pattern of electronic circuit traces.

CLAIM CHECK Artificial Intelligence

Hubinger’s AI Extinction Estimate Is a Personal Forecast, Not a Measured Result

Hubinger says present-model risk is low, but his posts supply no quantitative derivation for his greater-than-10-percent superintelligence forecast.

Shallow-focus close-up of multicolored JavaScript source code displayed on a computer monitor.

ANALYSIS Developer Tools

Visual Studio 18.10 Separates Model Choice From Its Preview Agent Harness

Visual Studio 18.10 adds provider choice to an early-preview agent, plus pull-request navigation and Podman debugger discovery.

Rows of black server racks display blue, green and red status lights in a computer room.

ANALYSIS Artificial Intelligence

SageMaker Feature Store Adds Atomic Partial Updates for Live ML Data

UpdateRecord can preserve unchanged features and avoid a preceding read, but AWS provides no measured savings or model-quality result.

A UML class diagram labeled Queue with one generic list field and create, enqueue and dequeue operations.

ANALYSIS Developer Tools

Kubernetes 1.37 Moves Group Scheduling to Beta for Coupled AI Workloads

Kubernetes 1.37 promotes group scheduling and preemption to Beta while leaving hierarchical workload placement behind Alpha gates.