AI news — February 18, 2026
The 12 top stories PulseAugur surfaced that day, ranked by signal across labs, papers, and developer communities.
-
Honest Ethics & AI – Part 1: The origins of morality
This multi-part essay sequence explores the origins of morality and its relation to artificial intelligence. The author argues that current AI systems, particularly transformer-based LLMs, are not equipped for moral decision-making due to their inherent lack of moral judgment. T…
-
Strava for Claude Code tracks AI developer usage and spend
A new tool called Strava for Claude Code has been released, designed to help developers track their usage and costs associated with AI models like Claude. The tool provides metrics on token consumption, iteration speed, and daily usage streaks, aiming to foster a competitive env…
-
Building Technology to Drive AI Governance
Researchers are developing new frameworks and tools to address the growing challenges in AI governance. One approach, the Agent Viability Framework, proposes an Informational Viability Principle for adaptive runtime governance of autonomous agents, focusing on estimating unobser…
-
Fei-Fei Li's World Labs secures $1B from A16Z, Nvidia for AI world models
Fei-Fei Li's AI startup, World Labs, has secured $1 billion in a new funding round. The investment was backed by major players including Autodesk, Andreessen Horowitz, Nvidia, and Advanced Micro Devices. This funding aims to advance the company's unique approach to developing AI.
-
Browser-based dev environment and AI skills pipeline automate coding tasks
A developer has created a self-contained Unix-like environment within a single 420KB HTML file, accessible in a browser without a server. This environment includes a shell, Git, Node.js, a C compiler, SQLite, Python, and integrates with the Claude Code API for AI-assisted coding…
-
IBM and UC Berkeley identify causes of enterprise AI agent failures
IBM Research and UC Berkeley have developed IT-Bench, a new benchmark designed to evaluate the performance of enterprise AI agents. They also introduced MAST, a framework for diagnosing the root causes of agent failures. This work aims to improve the reliability and effectivenes…
-
Google Gemini app adds AI music generation with Lyria 3 model
Google DeepMind has launched Lyria 3, a new generative music model, integrated into the Gemini app. Users can now create custom 30-second music tracks by describing a mood, genre, or even uploading a photo for inspiration. The model generates lyrics, offers control over musical …
-
Replit and Databricks partner to speed up enterprise data app development
Replit and Databricks have launched a new integration designed to streamline the creation and deployment of enterprise data applications. This partnership allows users to develop applications within Replit's collaborative environment and then deploy them directly to Databricks A…
-
Cognitive Synthesis and Neural Athletes
Deloitte's Chief Innovation Officer, Deborah Golden, discussed how AI is transforming leadership and organizational dynamics. She highlighted the increasing cognitive load on leaders and the need for empathy and anti-fragility in the evolving workplace. Golden emphasized Deloitt…
-
Gradio's gr.HTML enables one-shot web app creation from any web page
Gradio has introduced a new feature, gr.HTML, that allows developers to create web applications with a single HTML file. This simplifies the process of building and deploying interactive AI demos, enabling users to share their creations more easily. The feature aims to lower the…
-
OpenAI launches EVMbench to test AI agents' smart contract security skills
OpenAI has introduced EVMbench, a new benchmark designed to evaluate the capabilities of AI agents in detecting, patching, and exploiting vulnerabilities within smart contracts. This benchmark utilizes a curated set of 117 vulnerabilities from audits and aims to improve the secu…
-
METR details its robust security and confidentiality measures for AI model access
METR has detailed its comprehensive approach to safeguarding confidential information and non-public AI models. Their strategy involves a multi-layered system of policies, technical setups, and established norms to prevent leaks and mitigate insider threats. This includes assign…