AI Research: Latest AI News and Analysis

Track AI Research updates with focus on product moves, market signals, and high-impact developments.

Latest Articles

• AI labs are failing to keep their own systems in check — No AI company fully implements basic safety controls for its own internal AI systems, according to Guidelight's first independent scorecard — Anthropic and OpenAI earn C+, Google D+, xAI D−, and Meta an outright F on six core safety practices.
• The Download: AI’s self-improvement problem, and what’s driving the heat — A new study finds AI agents still can't conduct open-ended research, casting doubt on near-term recursive self-improvement — and separately, OpenAI paused model work after its Astra model hit a "critical" safety risk threshold.
• Jigsaw Jeeves: Building a Puzzle Assistant using Computer Vision — A Python tutorial from Towards Data Science walks through building a jigsaw puzzle assistant using OpenCV, NumPy, and SciPy — framing piece-to-board matching as a fragment-to-reference problem with direct applications in satellite tile stitching, forensic document reconstruction, and manufacturing assembly verification.
• IBM’s new modular architecture for cryogenic systems — IBM's new modular cryogenic architecture links quantum processors across connected cooling cells—each supporting 2,000+ qubits—using 'L-coupler' technology, clearing the physical infrastructure path toward fault-tolerant Starling system expected in 2029.
• Child-monitoring apps might need a reboot — Research increasingly shows that invasive child-monitoring apps — now a $1.57B market — can backfire by eroding trust and adolescent autonomy, as experts advocate a shift from surveillance toward graduated digital independence.
• PXDepth: Pixel-Space Modeling for Structure Preserving Monocular Depth Estimation — PXDepth decouples global scene context modeling (large-patch ViT) from pixel-level depth prediction (Context-Modulated Pixel Transformer blocks), preserving fine object boundaries and structures that standard ViT-based depth estimators lose through coarse tokenization — with code and model weights publicly released.
• When Agents Act on Web3: An Attack-Surface Survey of MCP, Skills, and Tool Calling — A survey of AI agent attack surfaces in Web3 finds that the share of MCP tools modifying external state rose from 27% to 65% — and current defenses stop fewer than 30% of attacks, while model-level safety refuses fewer than 3%. Four properties of blockchains (irreversibility, signing authority, continuous autonomy, sequence-level composition) turn normally recoverable agent failures into permanent losses.
• Inductively Scalable, Single-Step Neural Surrogates for Wave-Scattering Inverse Problems — UC Berkeley researchers train a neural surrogate for 2D wave-scattering that scales inductively to 3 million controllable variables — 73.8x beyond its training domain — while delivering speedups of up to 26.5x over FDTD simulation for photonic inverse design.
• ORPA: Online Residual Policy Adaptation for Robot Manipulation Control with Human Feedback — ORPA enables imitation-learned robot manipulation policies to recover from small errors in real time by predicting joint-space residual adjustments — without modifying the underlying policy — tested on the ALOHA platform.
• Multi-turn Conversational AI from Text to Multimodal Interaction: Data, Models, Evaluation, and Open Challenges — A comprehensive survey of multi-turn conversational AI finds that multimodal perception has advanced faster than contextual coherence — current systems across text, audio, and multimodal domains still fail on persistent memory, cross-turn grounding, and full-duplex interaction.