Category intelligence

AI News Briefing — February 16, 2026

9 current items analyzed and ranked.

Executive synthesis

AI News Summary

Google AI led the week's most impactful developments with WebMCP, a new protocol enabling structured website-to-AI-agent communication that could replace inefficient screen-scraping approaches. Separately, Google AI Overviews faced scrutiny for scam vulnerability in its search summaries.

The UK government announced planned legislation with fines or bans for AI chatbot makers endangering children, following the Grok/X image generation scandal
Guardian editorial flagged a wave of AI safety researcher departures from major labs, warning profit motives are overriding safety
Ars Technica retracted an article containing AI-fabricated quotations, underscoring risks of AI tool overreliance in journalism

In product and open-source releases, Moonshot AI launched Kimi Claw with 5,000+ community skills as a cloud-native agentic platform. Kani-TTS-2, a 400M parameter open-source TTS model running on 3GB VRAM with voice cloning, pushed the efficiency frontier. OpenClaw surpassed 150,000 GitHub stars as a self-hosted personal AI assistant.

Key Themes

AI Safety & Regulation · 3Agentic AI Infrastructure · 3AI Reliability & Trust · 2Open Source AI Models · 2

Primary evidence

Top Ranked Signals

News MarkTechPost Feb 15

Google AI Introduces the WebMCP to Enable Direct and Structured Website Interactions for New AI Agents

By Michal Sutter

82 score

AI Analysis

Google AI has introduced WebMCP (Web Model Context Protocol), a new standard that allows websites to communicate their available tools directly to AI agents, replacing the inefficient screen-scraping approach. This could fundamentally change how AI agents interact with the web by enabling structured, reliable browser-based automation.

Google is officially turning Chrome into a playground for AI agents. For years, AI ‘browsers’ have relied on a messy process: taking screenshots of websites, running them through vision models, and guessing where to click. This method is slow, breaks easily, and consumes massive amounts of compute. Google has introduced a better way: the Web Model Context Protocol (WebMCP). Announced alongside the Early Preview Program (EPP), this protocol allows websites to communicate directly t

Agentic AI InfrastructureWeb StandardsGoogle AI

News AI (artificial intelligence) | The Guardian Feb 15

Makers of AI chatbots that put children at risk face big fines or UK ban

By Robert Booth UK technology editor

78 score

AI Analysis

UK PM Keir Starmer will announce legislation imposing massive fines or service bans on AI chatbot makers whose products put children at risk. The move follows the Grok/X scandal involving AI-generated sexualized images of real people.

Starmer to announce ‘crackdown on vile illegal content created by AI’ after scandal involving Elon Musk’s Grok toolMakers of AI chatbots that put children at risk will face massive fines or even see their services blocked in the UK under law changes to be announced by Keir Starmer on Monday.Emboldened by Elon Musk’s X stopping its Grok AI tool from creating sexualised images of real people in the UK after public outrage last month, ministers are planning a “crackdown on vile illegal content crea

AI RegulationAI SafetyChild SafetyUK Policy

News AI (artificial intelligence) | The Guardian Feb 15

The Guardian view on AI: safety staff departures raise worries about industry pursuing profit at all costs | Editorial

By Editorial

72 score

AI Analysis

The Guardian editorial highlights a wave of notable AI safety researchers quitting major firms, warning that profit-chasing is sidelining safety work. The piece calls for urgent regulation before AI becomes 'too big to fail.'

Cash-hungry Silicon Valley firms are scrambling for revenue. Regulate them now before the tech becomes too big to failHardly a month passes without an AI grandee cautioning that the technology poses an existential threat to humanity. Many of these warnings might be hazy or naive. Others may be self-interested. Calm, level-headed scrutiny is needed. Some warnings, though, are worth taking seriously.Last week, some notable ground-level AI safety researchers quit, warning that firms chasing profits

AI SafetyIndustry EthicsAI RegulationTalent Movement

News MarkTechPost Feb 15

Moonshot AI Launches Kimi Claw: Native OpenClaw on Kimi.com with 5,000 Community Skills and 40GB Cloud Storage Now

By Michal Sutter

68 score

AI Analysis

Moonshot AI has launched Kimi Claw, bringing its OpenClaw agentic AI framework natively to kimi.com as a cloud-hosted platform. It features ClawHub with over 5,000 community-contributed skills and 40GB cloud storage for persistent AI agent environments.

Moonshot AI has officially brought the power of OpenClaw framework directly to the browser. The newly rebranded Kimi Claw is now native to kimi.com, providing developers and data scientists with a persistent, 24/7 AI agent environment. This update moves the project from a local setup to a cloud-native powerhouse. This means the infrastructure for complex agents is now fully managed and ready to scale. ClawHub: A Global Skill Registry The core of Kimi Claw’s versatility is ClawHub. This

Agentic AIChinese AIPlatform LaunchOpen Source

News Ars Technica - All content Feb 15

Editor’s Note: Retraction of article containing fabricated quotations

By Ken Fisher

67 score

AI Analysis

Ars Technica retracted an article after discovering it contained AI-fabricated quotations attributed to a real source. The outlet acknowledged this as a serious failure of editorial standards, despite having explicit policies against publishing unlabeled AI-generated content.

On Friday afternoon, Ars Technica published an article containing fabricated quotations generated by an AI tool and attributed to a source who did not say them. That is a serious failure of our standards. Direct quotations must always reflect what a source actually said. That this happened at Ars is especially distressing. We have covered the risks of overreliance on AI tools for years, and our written policy reflects those concerns. In this case, fabricated quotations were published in a manner

AI HallucinationsMedia EthicsAI Tool RisksJournalism

News MarkTechPost Feb 15

Meet ‘Kani-TTS-2’: A 400M Param Open Source Text-to-Speech Model that Runs in 3GB VRAM with Voice Cloning Support

By Michal Sutter

65 score

AI Analysis

As first reported on Reddit yesterday, Kani-TTS-2 is a new open-source 400M parameter text-to-speech model that runs on just 3GB VRAM and supports voice cloning. It uses an 'audio-as-language' approach with a neural codec, offering a lean alternative to closed-source TTS APIs.

The landscape of generative audio is shifting toward efficiency. A new open-source contender, Kani-TTS-2, has been released by the team at nineninesix.ai. This model marks a departure from heavy, compute-expensive TTS systems. Instead, it treats audio as a language, delivering high-fidelity speech synthesis with a remarkably small footprint. Kani-TTS-2 offers a lean, high-performance alternative to closed-source APIs. It is currently available on Hugging Face in both English (EN) and Portugue

Open Source AIText-to-SpeechEfficient AIVoice Cloning

News Feed: Artificial Intelligence Latest Feb 15

Google’s AI Overviews Can Scam You. Here’s How to Stay Safe

By David Nield

62 score

AI Analysis

Google's AI Overviews feature in search is being exploited by scammers who inject deliberately misleading information into AI-generated summaries, leading users to potentially harmful content.

Beyond mistakes or nonsense, deliberately bad information being injected into AI search summaries is leading people down potentially harmful paths.

AI SecuritySearch AIGoogleScams

News MarkTechPost Feb 15

Getting Started with OpenClaw and Connecting It with WhatsApp

By Arham Islam

55 score

AI Analysis

A tutorial on setting up OpenClaw, an open-source self-hosted AI assistant that has crossed 150,000 GitHub stars. It integrates with WhatsApp, Telegram, Slack, and 50+ services while keeping data under user control.

OpenClaw is a self-hosted personal AI assistant that runs on your own devices and communicates through the apps you already use—such as WhatsApp, Telegram, Slack, Discord, and more. It can answer questions, automate tasks, interact with your files and services, and even speak or listen on supported devices, all while keeping you in control of your data. Rather than being just another chatbot, OpenClaw acts as a true personal assistant that fits into your daily workflow. In just a few months,

Open Source AIPersonal AI AssistantsPrivacyTutorials

News AI (artificial intelligence) | The Guardian Feb 15

No swiping involved: the AI dating apps promising to find your soulmate

By Aisha Down

35 score

AI Analysis

New AI-powered dating apps use agentic AI to conduct personality interviews and provide curated matches based on 'similarity and reciprocity of personality,' eliminating traditional swiping mechanics.

Agentic AI apps first interview you and then give you limited matches selected for ‘similarity and reciprocity of personality’Dating apps exploit you, dating profiles lie to you, and sex is basically something old people used to do. You might as well consider it: can AI help you find love?For a handful of tech entrepreneurs and a few brave Londoners, the answer is “maybe”. Continue reading...

AI ApplicationsConsumer AIDating Tech