OpenAI announced GPT-5.4-Cyber, a variant of GPT-5.4 fine-tuned for defensive cybersecurity work. The model features binary reverse engineering capabilities and reduced safety restrictions, but access is limited to authenticated security professionals through the company's Trusted Access for Cyber program.
OpenAI released GPT-5.4-Cyber, a model variant designed for defensive cybersecurity tasks with fewer restrictions on dual-use queries. Access is controlled through a tiered verification system in the Trusted Access for Cyber program, targeting thousands of vetted users compared to Anthropic's 40-organization Mythos Preview rollout.
The UK's AI Safety Institute published an independent evaluation confirming Anthropic's Claude Mythos is highly effective at finding security vulnerabilities. The evaluation revealed a linear relationship: more tokens spent equals more exploits discovered, transforming security into an economic arms race.
Google Home's April 2026 update addresses Gemini voice assistant reliability issues. The update improves speech detection to reduce mid-sentence interruptions, speeds up responses to simple queries, and enhances music playlist recognition even when names are misspoken or in noisy environments.
Anthropic released Python SDK v0.95.0 on April 14, 2026, marking Claude 4 Sonnet and Opus models as deprecated. The update also includes Bedrock authentication improvements for AWS deployments.
Anthropic co-founder Jack Clark confirmed the company briefed the Trump administration on its Mythos model, which the company says is too dangerous for public release due to powerful cybersecurity capabilities. The briefing occurred despite Anthropic's ongoing lawsuit against the Department of Defense over AI system access restrictions.
Google is rolling out Skills, a Chrome feature that allows users to save frequently used AI prompts and reuse them with a single click. The feature works with Gemini across multiple tabs and is available now on Mac, Windows, and ChromeOS for English-US users.
The UK's AI Safety Institute reports Claude Mythos Preview achieved a 73% success rate on expert-level capture-the-flag cybersecurity challenges and became the first AI model to complete a full 32-step simulated corporate network takeover, succeeding in 3 out of 10 attempts. The testing occurred in environments without active security monitoring or defenders.
Google has released Skills for Gemini in Chrome on desktop, a feature that lets users save and execute frequently used prompts as one-click workflows. Users invoke Skills by typing a forward slash (/) in the prompt box, and the system can operate on the current page or multiple selected tabs.
Google is rolling out Skills, a new Chrome feature that lets users save Gemini prompts for one-click reuse. Available free to all Chrome users with US English language settings, Skills sync across devices and include a pre-built library of common prompts.
Anthropic has released routines for Claude Code, allowing automated bug fixes, code reviews, and deployment monitoring to run on cloud infrastructure without requiring users' local machines. The feature is available in research preview with daily limits of 5-25 runs depending on subscription tier.
Anthropic launched repeatable routines for Claude Code in research preview, allowing developers to schedule automated tasks that run on cloud infrastructure without requiring their local machine to be online. Pro users get 5 daily routines, Max users get 15, and Team/Enterprise users get 25.
Google is rolling out Personal Intelligence to Gemini users globally, excluding the European Economic Area, Switzerland, and UK. The opt-in feature accesses data from Gmail, Calendar, Drive, Google Photos, YouTube, Maps, and other first-party apps to provide contextual responses without additional prompting.
Google is adding Skills to Chrome, a feature that lets users save and reuse Gemini AI prompts across different webpages. The feature launches today for Chrome desktop users signed into Google accounts with English (US) language settings.
Google Chrome is launching Skills, a feature that lets users save Gemini AI prompts for instant reuse across browser tabs. The feature rolls out today for Chrome desktop users with US English language settings.
Anthropic is restricting access to Claude Mythos Preview, a model the company claims can find security vulnerabilities better than most humans, to 52 technology partners. While the UK's AI Security Institute has already tested Mythos and published findings, most European cybersecurity agencies have limited or no access, revealing a structural gap in Europe's AI safety evaluation capacity.
Anthropic's Claude.ai and Claude Code suffered a 48-minute outage on April 13, 2026, from 15:31 to 16:19 UTC with elevated error rates. GitHub quality complaints have increased 3.5× from the January-February baseline, though SWE-Bench-Pro scores show no substantive change since February.
An internal memo from OpenAI Chief Revenue Officer Denise Dresser outlines a new model codenamed 'Spud' for improved reasoning and production reliability, an agent platform called 'Frontier,' and expanded Amazon partnership. The memo also accuses Anthropic of inflating its $30B run rate by roughly $8B through gross revenue accounting.
Microsoft is testing OpenClaw-like features for Microsoft 365 Copilot aimed at enterprise customers, the company confirmed to The Information. The agent would run continuously to complete multi-step tasks over extended periods, distinguishing it from Microsoft's existing cloud-based agents like Copilot Cowork and Copilot Tasks.
Researchers have introduced LPM 1.0, an AI model that generates real-time video of a speaking, listening, or singing character from a single image, with lip-synced speech and facial expressions stable for up to 45 minutes. The system integrates directly with voice AI models like ChatGPT but remains a research project with no planned public release.