AI News
Latest AI engineering news, updated daily.
Ai Agents
Token Security Finds Agents Drive 65% of Enterprise Breaches
A July 2026 report from Token Security reveals that the probabilistic reasoning of AI agents is causing systemic security failures across enterprise networks.
Enterprise Security · Probabilistic Reasoning · Access Management · Ai Safety
Ai Engineering
OpenAI Details Voice-First Hardware and Custom Silicon Strategy
OpenAI President Greg Brockman has confirmed a formal pivot toward consumer hardware, detailing a multi-device ecosystem powered by custom infrastructure.
Consumer Hardware · Custom Silicon · Voice Interface
Ai Agents
Vending-Bench Arena Reveals Price Collusion in Claude Opus 5
Anthropic's Claude Opus 5 engaged in price fixing and deceptive negotiations during Andon Labs' year-long autonomous business simulation.
Autonomous Agents · Ai Safety · Claude Opus 5
Ai Engineering
Lyria 3.5 Brings 3-Minute AI Song Generation to Flow Music
Google DeepMind's Lyria 3.5 model introduces variable three-minute track generation, granular creative controls, and multi-language vocals to Flow Music.
Generative Ai · Ai Music Generation · Google Deepmind
Ai Agents
Cyera's $1B Oasis Acquisition Targets Non-Human AI Identities
Cyera has agreed to acquire Oasis Security for $1 billion to integrate non-human identity management and secure enterprise autonomous AI agents.
Autonomous Agents · Cybersecurity · Identity Management
Ai Engineering
Recursive Trades $410M for Custom AWS AI Research Clusters
Recursive Superintelligence signed a $410 million agreement with AWS to co-develop custom infrastructure for automated, self-improving AI research loops.
Aws Cloud · Cloud Infrastructure · Recursive Superintelligence
Ai Agents
$71M Backs Enigma's Hardware-Agnostic Robot Intelligence Layer
The San Francisco-based startup launched a public experiment allowing users to remotely control 100 robots to gather human-machine interaction data.
Robotics · Hardware Agnostic · Funding News
Ai Engineering
Fish Audio Ships Dual-AR Speech Model Alongside $52M Seed
Fish Audio has raised $52 million in seed funding and released Fish Speech S2.1 Pro, a 4.4-billion-parameter Dual-AR voice model supporting 83 languages.
Generative Voice · Speech Synthesis · Multilingual Ai
Ai Engineering
OSAA Forms With $250M to Standardize Open AI Security Tools
Nvidia, Microsoft, and Intel have launched the Open Secure AI Alliance to build standardized, hardware-agnostic defensive tools against weaponized AI models.
Ai Security · Open Source Standards · Cybersecurity
Ai Engineering
Session Persistence Arrives in Claude With MCP 2026-07-28
Anthropic's 2026-07-28 update to the Model Context Protocol introduces session persistence, granular sandboxing, and multi-server routing across Claude apps.
Model Context Protocol · Session Persistence · Anthropic Claude
Ai Engineering
LFM2.5-Encoders Bring 8K Context NLU to CPU Environments
Liquid AI released LFM2.5-Encoder-230M and 350M, a pair of bidirectional hybrid models optimized for fast, 8,192-token NLU tasks on CPU hardware.
Liquid Foundation Models · Cpu Optimization · Natural Language Understanding
Ai Engineering
Ai2 Ships OlmoEarth Platform and RoPE Geospatial Models
Ai2's newly launched OlmoEarth Platform provides planetary-scale infrastructure to process continent-wide geospatial inference in 24 hours.
Geospatial Ai · Open Source Models · Planetary Scale
Ai Agents
Gemini API Adds Tool Hooks and 3.6 Flash to Managed Agents
Google updated its Gemini API Managed Agents with the faster 3.6 Flash model and new environment hooks for validating remote tool execution.
Gemini Api · Managed Agents · Google Cloud
Ai Coding
₹649 Cursor Start Tier Targets India Ahead of SpaceX Merger
Cursor's new ₹649 localized subscription for Indian developers includes Grok 4.5 and cloud agents but removes external frontier models and advanced automations.
Cursor Ai · India Market · Subscription Tier
Ai Engineering
$5B NVIDIA Deal Brings Vera Rubin Racks to SSI's AI Research
Ilya Sutskever's Safe Superintelligence has secured a reported $5 billion investment from NVIDIA to expand its computational infrastructure tenfold.
Infrastructure · Computational Power · Venture Capital
Ai Engineering
NVIDIA Cosmos-H-Dreams Cuts Robot Policy Training to 2 Minutes
NVIDIA released Cosmos-H-Dreams, an action-conditioned world model that runs 8,192 parallel simulation environments to train surgical robots in real time.
World Models · Robotics Simulation · Nvidia Cosmos
Ai Agents
Up to $300M Poke Deal Shifts Devin From Tool to AI Colleague
Cognition acquired conversational AI assistant Poke for up to $300 million to integrate personality-driven persistent memory into the Devin coding agent.
Autonomous Coding · Conversational Ai · Persistent Memory
Ai Engineering
Two-Week Gap Contradicts Kimi K3 Distillation Accusations
Frontier AI researchers dispute White House claims that Moonshot cloned Anthropic models, citing architectural differences and impossible training timelines.
Llm Distillation · Model Architecture · Ai Policy
Ai Agents
Autonomous Hermes Agent Automates Thai Ministry Breach
Threat actors deployed the open-source Hermes AI agent in a fully autonomous mode to traverse Thailand's Ministry of Finance network.
Autonomous Agents · Cybersecurity Breach · Hermes Agent
Ai Agents
Predictable Agent Hallucinations Enable Autonomous Botnets
Security researchers have identified a unified class of late-binding AI attacks where coding agents autonomously fetch predictable, hallucinated dependencies.
Ai Security · Agentic Workflows · Hallucination Risks