The Verge

Iranian team Explosive Media creates viral Lego war videos with AI
Iranian group Explosive Media is getting millions of views with Lego-style AI videos mocking US military failures in Iran, and explains the

AI companion in a plush deer wrote an unprompted theory about Mitski’s father’s CIA ties
A toy with the Coral AI companion sent a user a conspiracy theory about singer Mitski without any prompt — and it turned out to be fabricate

The AI coding race: how OpenAI, Google, and Anthropic are competing for the developer market
The war over the market for AI tools for programmers has flared up between OpenAI, Google, and Anthropic — and it all began with GitHub Copi

Zuckerberg Creates His Own AI Clone for Employee Meetings
Meta trains a digital clone of Zuckerberg using his voice, appearance, and manner of speech to replace the CEO at team meetings.

Microsoft Explores Integration of OpenClaw-Agents into 365 Copilot for Enterprise Markets
Microsoft Tests OpenClaw-like Agents for 365 Copilot, Aiming to Make the AI Assistant Fully Autonomous — to Execute Tasks for Users Round th

Google Chrome launched “Skills” — saved AI commands for Gemini
Chrome now allows any prompt for Gemini to be saved as a “Skill” and launched on any page with one click.

Attacks on Sam Altman: how violence became a weapon against the AI industry
A Molotov cocktail at the home of the OpenAI CEO and gunfire at a city official’s door — the wave of violence against the AI sector is growi

Apple secretly threatened to remove Grok from the App Store over deepfakes on X
In January, Apple demanded in a private letter that the X and Grok teams develop a content moderation plan after a wave of sexual deepfakes.

Adobe announces Firefly AI Assistant — editing through description instead of manual tools
Adobe introduced Firefly AI Assistant — an AI assistant for editing through text descriptions, without needing knowledge of specialized Crea

Google releases Gemini app for Mac with quick access via Option + Space
Google has launched a native Gemini app for macOS: a pop-up chat via keyboard shortcut, screen access, and media generation right from the d

Canva AI 2.0: all design through a single text prompt
Canva has updated the platform to AI 2.0 — now all tools are available through a unified chat interface with natural-language prompts.

Character.AI Launches Books Mode: Roleplaying Games Based on Classic Literature
Character.AI launches Books Mode with 20+ classics in public domain—an attempt to reposition roleplay chats after wave of scandals

Anthropic Releases Claude Opus 4.7 — New Flagship for Complex Coding
Anthropic introduced Claude Opus 4.7 — the company's most powerful publicly available model, featuring improved coding, image analysis, and

Gemini now creates personalized images from your Google Photos
Google expanded Personal Intelligence: Gemini uses your photos from Google Photos and the Nano Banana 2 model to generate images tailored to

OpenAI Updated Codex: Desktop Management, Image Generation, and Task Memory
OpenAI has expanded Codex's capabilities: the agent can now manage applications on the macOS desktop, generate images, and retain context fr

Google updated AI Mode: sources now open next to the chat
Google added a source-viewing mode to AI Mode without leaving the chat — the website opens on the side, and you can immediately ask question

The “AI Is Inevitable” Trap: Have We Reached Peak Hype Around Artificial Intelligence
Footwear maker Allbirds declared itself an AI company and saw its value rise sevenfold in a day — an episode that became a symbol of the cur

Dairy Queen rolls out voice AI at dozens of drive-thrus across the US and Canada
Ice cream chain Dairy Queen is launching Presto's AI chatbot at drive-thrus to speed up service and increase average order value.

Anthropic reconciles with Trump: new Claude Mythos model for cybersecurity
After two months of public conflict, Anthropic may restore relations with the Trump administration thanks to the specialized Claude Mythos P

Sora Team Lead Bill Pebbles Leaves OpenAI After Project Shutdown
After OpenAI shut down its Sora video generator, team lead Bill Pebbles announced his departure from the company.

Epic Games makes AI character creation tool available to Fortnite developers
Epic Games launched Conversations, a tool that lets map creators build NPCs with unscripted AI-based dialogue.

Yelp Transforms AI Assistant Into a Digital Concierge With Booking
Yelp Assistant now sits at the center of the app: recommends restaurants, answers questions and books tables — all in one conversation witho

Джон Тернус сменит Тима Кука в Apple — и сразу столкнётся с AI-провалом
Apple объявила: с 1 сентября CEO станет Джон Тернус — 25-летний ветеран из hardware. Его главный вызов — AI, который у Apple пока безнадёжно

YouTube расширяет защиту от дипфейков: голливудские звёзды смогут удалять ИИ-клоны
YouTube запустил инструмент обнаружения дипфейков для знаменитостей — они смогут отслеживать ИИ-видео с собой и запрашивать их удаление с пл

«The AI Doc»: почему новый фильм об ИИ страдает от избыточного хайпа
Разбираемся, почему документальный фильм «The AI Doc» не смог стать объективным взглядом на индустрию, утонув в крайностях между предсказани

Очки Meta с ИИ отправляют интимные видео на проверку людям в Кении
Расследование шведских журналистов выявило, что подрядчики Meta в Найроби просматривают видео с умных очков, включая записи из ванных комнат

ИИ научился раскрывать анонимные аккаунты в соцсетях
Исследователи из ETH Zurich и Anthropic создали систему ИИ-агентов, способную связывать анонимные аккаунты с реальными людьми. Эпоха безопас

Google встраивает рабочее пространство Canvas прямо в поисковую выдачу
Google открывает Canvas для всех пользователей AI Mode в поиске. Теперь прямо из поисковой строки можно создавать документы, писать код и пл

Raycast запускает Glaze — платформу, где вайб-кодинг становится по-настоящему простым
Создатели популярного Mac-лаунчера Raycast выпустили Glaze — приложение, которое убирает последние барьеры между обычными пользователями и с

Gemini научился заказывать еду и такси вместо вас
Google выпустила мартовское обновление Pixel с агентными функциями Gemini: ИИ-ассистент теперь может самостоятельно выполнять задачи в прило