All coverage

The latest in technology, clearly sourced.

One stream across AI, mobile, security, startups, enterprise technology, and gaming.

News stream

Page 70
1243
AI

Meta Launches Muse Image AI Model With Instagram User Tagging Across Its Apps

Meta’s Superintelligence Labs launched Muse Image in July 2026 for Meta AI, Instagram, and WhatsApp. It lets users @ mention Instagram accounts in prompts so Meta AI can use public photos for generated likenesses (with controls), plus transforms and edits—starting with 30 new Instagram Stories AI effects in the US.

The Verge
1245
General

X Launches In-App Video Editor as It Acknowledges Widespread Video Theft by Top Accounts

X added an in-app video recorder/editor to its iOS app in July 2026, after product head Nikita Bier said many videos from top accounts are “simply stolen,” sometimes 5 years after going viral, and make up close to half of impressions. The update brings trimming, multi-language auto-captions, and green-screen backgrounds from camera-roll photos, as X tackles content recycling by redirecting impressions from programmatic reuploads and cutting payouts for stolen reposts/clickbait.

The Verge
1246
AI

Figma Acquires Team Behind Vibe-Coding and AI Agent Platform Bud

Figma’s July 2026 acquisition of Bud (formerly Orchids) goes beyond expansion into app prototyping: both platforms will shut down July 18, 2026, and users must migrate. It also follows BBC reporting earlier this year on Orchids-made apps’ cyberattack exposure.

TechCrunch
1247
AI

Microsoft Shifts to In-House AI Models to Cut Costs in Excel and Word

Microsoft is moving some Excel and Word prompts from OpenAI/Anthropic to its in-house MAI models, Bloomberg reported July 7, 2026, citing rising AI service costs. At last month’s Build, Microsoft unveiled seven new MAI models, including an agentic coder and text-to-image generator, while Amazon, Uber, Meta, and Accenture cut AI spend too.

TechCrunch
1250
Mobile & Apps

Bidbus Raises $15 Million to Let Dealers Bid Against Each Other for Used Cars

Bidbus raised $15M in a July 2026 Series A to expand its real-time used-car auction where dealers get only a few hours to submit competing bids, with offers averaging $2,000–$3,000 more than Carvana even after Bidbus’s cut. Led by Ibex Investors with Mucker Capital, FJ Labs and others, the Los Angeles startup previously operated only in California and Texas.

TechCrunch
1251
AI

AI Law Startup Norm Raises $120M Series C, Reaches $1.2B Valuation

On July 7, 2026, Norm raised a $120M Series C led by Khosla Ventures, valuing the nearly three-year-old AI law firm at $1.2B. Norm Law uses human-supervised AI agents for enterprise work and charges by outcomes, not hourly. Other investors include Bain and Fenwick LLP.

TechCrunch
1253
Mobile & Apps

X Launches Video Editor for iOS to Push Creators Toward Original Content

X links its recycled video problem to UX and business impact, saying videos account for nearly half of all impressions. On July 6, 2026, product lead Nikita Bier launched an iOS editor/recorder with multilingual caption overlays, customizable caption styles, and green-screen background swaps; Android is still being rebuilt.

TechCrunch
1257
AI

Apple Adds Siri Voice Pace and Expressivity Controls in iOS 27 Beta 3

Apple’s iOS 27 developer beta 3 (July 6, 2026) adds Siri “Pace” and “Expressivity” sliders to adjust how quickly it speaks and how much human-like emotion its voice conveys; they were first announced at WWDC 26 in June. The generative-AI Siri in iOS 27 is integrated for voice, Dynamic Island swipes, typing, and a stand-alone Siri app.

TechCrunch
1259
AI

Google Is Using Your Uploaded Media to Train AI — Here’s How to Opt Out

Google’s June 2026 privacy update automatically opts users into letting uploaded media (images, audio, video, files) be stored and used to train AI after an email notice. It now covers services like Maps, Shopping, Flights, Hotels, Translate and News, and moves the opt-out from Web & App Activity to a new Search Services History setting.

TechCrunch
1260
AI

Reddit Uses AI to Fight AI-Generated Spam, Cutting User Exposure by 20%

Reddit says its 2026 LLM-powered anti-spam tools block 23M spam views a day and flag ~25,000 new spam posts and comments daily. It cut user exposure to spam by 20% from Jan–Mar 2026 vs. the prior three months; YouTube, Meta and Instagram require disclosure, and TikTok is testing feed controls for AI content.

TechCrunch