2026

Transformers at the wheel: how Yandex uses AI to control self-driving vehicles

Google DeepMind's Gemma 4 arrives on Amazon Bedrock: three models with MoE and multimodality

AI Red Teaming — How Companies Test AI Systems for Vulnerabilities Before Release

AI red teaming is a method of testing AI systems through simulated attacks. It helps identify vulnerabilities before launch and meet regulat

01AWS Strands Evals automatically identifies root causes of AI agent failures and suggests fixesJun 30, 2026·3 min 02Csquare Files for IPO in US: Data Center Backed by Brookfield Goes PublicJun 30, 2026·2 min 03NVIDIA BioNeMo lets you fine-tune biological AI models with LoRA in hoursJun 30, 2026·2 min 04Amazon Bedrock Guardrails Gets a New API for Flexible Protection of Agentic AI ApplicationsJun 30, 2026·2 min 05LangSmith Added End-to-End OpenTelemetry Support for LangChain and LangGraphJun 30, 2026·2 min

Top themes

LLM·MIT Technology Review

Data centers connect to the grid faster through flexible electricity consumption agreements

A new flex-connection strategy allows data centers to avoid multi-year waiting queues: they agree to reduce consumption during peak hours in

01LangChain released Promptim, a library for automatic prompt optimizationJun 30, 2026·2 min 02Sarah Guo of Conviction: why LLM won't replace work that can't be measuredJun 30, 2026·2 min 03New Computer Raised Agent Memory Search Accuracy by 50% with LangSmithJun 30, 2026·2 min

LLM·3DNews AI

Qualcomm Develops Over 40 AI Devices and Predicts the End of the App Era

Qualcomm CEO Cristiano Amon believes AI agents will render traditional applications unnecessary — the company is already developing around 4

LLM·Habr AI

Why AI pilots don’t make it to production: what must be in place before scaling

A successful AI pilot does not guarantee a production rollout. Six governance controls without which any AI solution will remain stuck at th

LLM·LangChain Blog

LangChain Launches Headless Tools: Agents Work Directly in the Browser

LangChain released headless tools that enable AI agents to safely execute tasks on the client side — with access to browser APIs and local data without transmission to the server.

LLM·Ars Technica

Vulnerability in Microsoft Copilot Allowed Stealing Two-Factor Authentication Codes

LLM·TNW

YouTube deleted 16 channels with 35 million subscribers in AI-slop purge — and caught honest creators

LLM·Guardian

Anthropic Fable classified as a munition: the US blocked the AI model worldwide

LLM·Machine Learning Mastery

Scikit-LLM: multi-label text classification without a training dataset using LLMs

LLM·Habr AI

AI assesses Russians' job suitability from their faces — developers move into the public sector

Meta explains the difference between GPU, CPU, and its own MTIA chips in AI infrastructure

How to speed up transformer training with NVIDIA Apex: FusedAdam, FusedLayerNorm, and torch.amp

NVIDIA speeds up LLM pretraining: NVFP4 on Blackwell paired with JAX and MaxText

LLM·AWS Machine Learning Blog

AWS launched OAuth authorization for MCP servers in Bedrock AgentCore Gateway

NVIDIA JetPack 7.2: AI agents on edge devices with a single deployment command

84% of Russian mobile apps contain critical vulnerabilities — and their number is growing

LLM·NVIDIA Developer Blog

NVIDIA TensorRT turns FP8 checkpoints into engines for fast inference

NVIDIA explained how to convert an FP8-quantized checkpoint into a TensorRT engine — a step that turns an optimized model into a real produc

01JetBrains released Mellum2 — an open 12B MoE model for fast tasks in AI pipelinesJun 30, 2026·2 min 02NVIDIA: how to design battery energy storage systems for AI factoriesJun 30, 2026·2 min 03Claude Fable 5 arrives on GitLab Duo — multi-day agent tasks without manual stopsJun 30, 2026·3 min 04GitLab launched a managed DevSecOps platform on Google Cloud with Gemini 3.5 FlashJun 30, 2026·2 min 05Alibaba launched Qwen3.7-Plus: multimodal agent with video analysis and autonomous codingJun 30, 2026·2 min

Top themes

LLM·Guardian

London's Royal Opera House launches RBO/SHIFT: artists in dialogue with AI instead of apocalyptic rhetoric

The RBO/SHIFT festival at London's Royal Opera House will bring together artists and technologists to honestly explore AI's role in art — wi

01LangChain published a guide to creating a custom harness for an AI agentJun 30, 2026·2 min 02LangGraph gets fault tolerance: retries, timeouts, and error handlers for production agentsJun 30, 2026·2 min 03OpenAI launches Codex: the AI agent that writes code instead of engineersJun 30, 2026·1 min

LLM·@demishassabis

Google released Gemma 4 12B: 150 million downloads, runs on a laptop with 16 GB of VRAM

Google DeepMind has released Gemma 4 12B — a compact open model that runs locally with 16 GB of VRAM, marking 150 million downloads for the

LLM·LangChain Blog

LangChain: why model neutrality protects businesses from vendor lock-in

LangChain warns: AI labs are building dependency not into the models, but into the tooling layer around agents — and only a neutral open-sou

LLM·Together AI Blog

Together AI beat TensorRT-LLM by 31% in benchmarks for code agents

Together Inference Engine delivered 31% more tokens per second and halved TTFT under peak load — the first fair test for production agents.

Replit: describe an idea in words — get a complete app with a database and deployment

LLM·@Replit

Replit launched parallel agents: an app, a video, and a pitch deck at the same time

Gopuff launched Grok-based AI agent Go — it handles voice, photos, and text

Together AI received ISO 27001:2022 certification for enterprise AI workloads

LLM·@demishassabis

Google DeepMind releases DiffusionGemma — a diffusion-based LLM 4x faster than other Gemma 4 models

Together AI: GPT-5.5, Gemini and Opus cannot write fast multi-GPU kernels

eToro launched Tori AI agent powered by Grok models for market sentiment analysis

LLM·@GoogleDeepMind

Google DeepMind launches a $10 million fund to study the collective behavior of AI agents

Google DeepMind and its partners have created a $10 million research fund to understand how millions of AI agents interact in groups and what unpredictable collective effects emerge from that.