請同意我們的隱私權規範，才能啟用聽新聞的功能。

（中央社記者張欣瑜舊金山19日專電）Google I/O年度開發者大會今天登場，發表新一代AI模型Gemini 3.5系列，在一般用戶與企業應用兩大市場迎戰OpenAI與Anthropic。Google強調，這一系列專為執行「代理任務」而優化。第一款模型3.5 Flash今天起開放給全球用戶使用。

Google I/O年度開發者大會19日登場，發表新一代AI模型Gemini 3.5系列。（Google提供）中央社記者張欣瑜舊金山傳真 115年5月20日

這意味著一般用戶使用Gemini App及Google搜尋的AI模式時，3.5 Flash作為背後預設模型。

Google也發表可被視為影片版Nano Banana的多模態影片生成與編輯模型Gemini Omni。

3.5 Flash今起開放 3.5 Pro預計6月推出

針對今年I/O核心方向，Google執行長皮查伊（Sundar Pichai）強調，人工智慧（AI）發展進入一個快速推進的階段，比起技術展示，使用者現在更加重視AI在產品當中的實際價值。

新一代3.5 Flash模型結合了「前沿智慧」與「行動能力」。後者是「AI代理」的關鍵核心。

一般用戶外，開發者可透過Google開發平台Antigravity，以及Google AI Studio、Android Studio的Gemini API使用。另外，3.5 Pro已在Google內部使用，預計6月對外推出。

3.5 Flash具備速度和成本優勢；Google表示，以每秒輸出的詞元（Token）數量計算，它的速度比其他前沿模型快約4倍；成本不到前沿模型一半，在某些情況下甚至可以降到約1/3。

另外，在幾乎各類型評測中，3.5 Flash的表現都超越上一代旗艦模型Gemini 3.1 Pro。

Omni多模態生成模型 想法可變電影感影片

影片來源：Google

除了主力模型，Google在今年I/O大會上還發表了Gemini Omni多模態生成模型，並指出可以把Omni想成圖像模型Nano Banana，只是它主要是用來製作影片。

現階段，Google先聚焦在影片生成編輯，最終，希望將Omni打造成一個世界模型，能夠接收圖片、聲音、影片與文字，並運用理解與推理能力，產生包含影片在內的多種輸出格式。

透過 Google News 追蹤中央社

Gemini Omni系列首款模型Omni Flash即日起透過Gemini App與Google Flow，開放給全球 Google AI Plus、Pro和Ultra訂閱用戶使用；同時也將從本週起，免費提供給YouTube Shorts與YouTube Create App使用者。至於下一款進階版本Omni Pro，預計在不久之後推出。

Google並強調，透過Omni生成或編輯的內容，都會自動加入SynthID數位浮水印，讓使用者能清楚辨識哪些是AI生成或經過AI工具編輯的內容，強化整體內容的透明度與安全性。（編輯：韋樞）1150520

請同意我們的隱私權規範，才能啟用聽新聞的功能。

（中央社記者張欣瑜舊金山19日專電）Google I/O年度開發者大會今天登場，發表新一代AI模型Gemini 3.5系列，在一般用戶與企業應用兩大市場迎戰OpenAI與Anthropic。Google強調，這一系列專為執行「代理任務」而優化。第一款模型3.5 Flash今天起開放給全球用戶使用。

這意味著一般用戶使用Gemini App及Google搜尋的AI模式時，3.5 Flash作為背後預設模型。

Google也發表可被視為影片版Nano Banana的多模態影片生成與編輯模型Gemini Omni。

● 3.5 Flash今起開放 3.5 Pro預計6月推出

針對今年I/O核心方向，Google執行長皮查伊（Sundar Pichai）強調，人工智慧（AI）發展進入一個快速推進的階段，比起技術展示，使用者現在更加重視AI在產品當中的實際價值。

新一代3.5 Flash模型結合了「前沿智慧」與「行動能力」。後者是「AI代理」的關鍵核心。

一般用戶外，開發者可透過Google開發平台Antigravity，以及Google AI Studio、Android Studio的Gemini API使用。另外，3.5 Pro已在Google內部使用，預計6月對外推出。

3.5 Flash具備速度和成本優勢；Google表示，以每秒輸出的詞元（Token）數量計算，它的速度比其他前沿模型快約4倍；成本不到前沿模型一半，在某些情況下甚至可以降到約1/3。

另外，在幾乎各類型評測中，3.5 Flash的表現都超越上一代旗艦模型Gemini 3.1 Pro。

● Omni多模態生成模型 想法可變電影感影片

除了主力模型，Google在今年I/O大會上還發表了Gemini Omni多模態生成模型，並指出可以把Omni想成圖像模型Nano Banana，只是它主要是用來製作影片。

現階段，Google先聚焦在影片生成編輯，最終，希望將Omni打造成一個世界模型，能夠接收圖片、聲音、影片與文字，並運用理解與推理能力，產生包含影片在內的多種輸出格式。

透過 Google News 追蹤中央社

Gemini Omni系列首款模型Omni Flash即日起透過Gemini App與Google Flow，開放給全球 Google AI Plus、Pro和Ultra訂閱用戶使用；同時也將從本週起，免費提供給YouTube Shorts與YouTube Create App使用者。至於下一款進階版本Omni Pro，預計在不久之後推出。

Google並強調，透過Omni生成或編輯的內容，都會自動加入SynthID數位浮水印，讓使用者能清楚辨識哪些是AI生成或經過AI工具編輯的內容，強化整體內容的透明度與安全性。（編輯：韋樞）1150520

請同意我們的隱私權規範，才能啟用聽新聞的功能。

（中央社記者張欣瑜舊金山19日專電）過去與Gemini對話，多半是使用者提問、Gemini回答，它的角色較像AI助理；Google今天在I/O年度開發者大會發布Gemini Spark，讓Gemini未來不再只是回應問題，而是能依照使用者指示，24小時自動追蹤資訊、處理實際任務。

Google今天推出Gemini Spark，並將它稱之為全天候個人AI代理（Personal AI Agent）。

這項應用運行在Google雲端伺服器上，因此即使用戶關閉筆電或手機，任務仍可持續執行。

根據Google Labs、Gemini暨AI Studio副總裁伍德華（Josh Woodward）介紹，早期測試者已使用Gemini Spark處理多種任務。例如，員工若需要寄送工作進度更新給主管，Spark可從電子郵件、文件、試算表和簡報中整理相關資訊，並協助撰寫草稿。

甚至有小型企業主讓它協助留意收件匣，確保不會漏掉潛在客戶的任何詢問。

在個人使用方面，Gemini Spark可以自動分析每月信用卡帳單，找出新增或容易被忽略的訂閱費用；也能整理Gmail信箱中孩子學校寄來的各種通知，轉成清楚摘要。

Google強調，AI代理人會依照使用者指示運作；使用者可自行決定是否啟用，以及要讓它連接哪些應用程式；在安全機制方面，它在執行高風險動作前會先徵求同意，例如付款。

Google正加速布局代理式AI時代，從AI晶片、雲端平台到Gemini產品都持續升級。Google上個月揭曉新一代TPU晶片（Tensor Processing Unit，張量處理單元），今天在I/O年度開發者大會發布Gemini 3.5模型家族，並展示Gemini如何從聊天式AI助理，走向代理式AI應用。

Google表示，Gemini Spark將率先提供給受信任測試者，並計畫下週向美國Google AI Ultra訂閱用戶開放Beta版。

除了Spark，Google也在搜尋中推出「資訊代理」（Information Agents）。

這個代理會搜尋網路上的各類資訊，包括部落格、新聞網站、社群貼文，以及金融、購物、體育等即時資訊。此功能將於今年夏天率先向Google AI Pro和Ultra訂閱用戶推出。（編輯：唐佩君）1150520

Google announced at its annual Google I/O event on Tuesday that it’s launching Pics, a new AI-powered design and image-generation app for Google Workspace. The tech giant says it designed the app to be accessible to everyone, from teachers to small business owners.

With Pics, users can generate everything from social media graphics and invitations to marketing materials and mock-ups using simple text prompts, without needing any editing skills or advanced tools. By giving users an easy way to generate visuals, Google is looking to take on popular design apps like Canva, as well as products from AI-native competitors like Claude Design from Anthropic. Google’s entry into the space signals that AI-powered design is fast becoming a core competitive arena — with real stakes for any business that depends on visual content.

The new app is launching to a group of testers at I/O and will be rolling out to Google AI Ultra subscribers this summer, Google says.

The company acknowledges that although AI models today can generate high-quality images, it’s still difficult to modify just one part of an image. If you get an image that’s almost perfect but want to change a small detail, you have to write an entirely new prompt and hope the AI doesn’t alter too much. That’s why Pics not only generates images but also makes them easily editable.

Loading the player…

Users can enter a prompt, and Pics will generate what they need. Gemini powers the editing layer, making every element in a generated design or image fully adjustable. You can write a new prompt to make changes, but you can also simply click the part you want to change and leave a comment — much like leaving feedback in Google Docs.

You can also edit directly, without leaving a comment or writing a prompt. For example, if you create a birthday party invitation and want to change the time listed on the card, you can do so manually.

Pics is powered by Nano Banana 2, which Google says is a strong fit for the app because it supports precise text rendering, real-world knowledge, and detailed visual output. Pics is also built natively into Google Workspace, enabling visual collaboration across its apps.

Once you’re happy with your design, you can download, copy, print, or share it with others. You can also pass it to someone else for a final round of edits before it goes out, Google says.

Catch up on the rest of Google IO 2026’s big news

Google Search as you know it is over

Google updates Gemini app to take on ChatGPT and Claude

Google introduces Gemini Spark, a 24/7 agent assistant with Gmail integration

How to use Google’s new information agents

Google isn’t finished infusing AI into your inbox. On Tuesday at their IO 2026 developer conference, the tech giant announced an expansion of its “AI Inbox” functionality for Gmail, which is adding conversational AI features. That means you can ask Gmail about things in your inbox instead of typing in search terms.

The company says the Gemini AI-powered feature, called Gmail Live, will help you quickly find information buried in your inbox.

Image Credits:Google

Perhaps you need information about your upcoming flight, the time of your dentist appointment, the door code for your Airbnb rental, or some details about an event at your kid’s school, for instance.

Before, you’d have to type in keywords in the search box (or maybe type in someone’s email address or domain) to try to narrow down your search. That doesn’t always make emails easy to find, however, especially if the search term is something found across several messages.

“Gmail Live can answer naturally phrased questions, respond to follow-up questions, and pivot if you need to interrupt it,” Devanshi Bhandari, product lead for Gmail, explained in a briefing ahead of Google’s annual developer conference, Google I/O, where the feature was first introduced to the public.

It’s another way that Google is trying to showcase how its AI technology can drive real-world improvements to products used by millions of consumers, at a time when many outside the tech industry are questioning the value of AI, as new data centers get built in their backyards, driving up their power bills.

Being able to point to something as simple as making it easier to find something that’s lost in your email inbox — an experience nearly everyone has suffered at some point — could be a practical and positive use case for AI … or at least, Google hopes.

Bhandari demonstrated Gmail Live to reporters, asking the tool a series of questions about things in the inbox, like a child’s show-and-tell project and their class trip, plus hotel and flight information for a trip to Detroit. Similar to using a stand-alone AI chatbot like Gemini or ChatGPT, Gmail users can ask these questions aloud in natural language, and the chatbot responds.

In the demo, Gmail Live also understood nuances between things like “field trip” and “trip” and was able to jump from one topic to another, Bhandari pointed out. Plus, the AI can pull granular details from emails, like a hotel room number, or infer which people you’re asking about, even when they’re not explicitly named.

Similar voice technology is also coming to its to-do list, Google Keep, the company noted.

Notably, Gmail Live is not replacing traditional Gmail search — it’s just another option.

Google may have learned that not everyone is ready for an AI-only experience after it “upgraded” Google Photos with AI-powered search to much backlash. Google Photos later rolled back the feature, making the use of AI optional after numerous complaints.

Gmail is also gaining other new capabilities, including ready-to-send drafts, instant file access, and the ability to manage to-dos by marking individual tasks as done.

Image Credits:Google

Plus, the AI Inbox experience, which launched earlier this year, will expand beyond Google AI Ultra subscribers to reach Google AI Pro and Plus subscribers as well. This allows you to see an overview of the tasks and items to catch up on that are buried in your inbox, all on one page.

The voice-powered Gmail Live feature, however, will roll out later this summer and will initially be limited to Google AI Ultra subscribers.

Catch up on the rest of Google IO 2026’s big news

Google Search as you know it is over

Google updates Gemini app to take on ChatGPT and Claude

Google introduces Gemini Spark, a 24/7 agent assistant with Gmail integration

How to use Google’s new information agents

At the Google I/O 2026 keynote, the tech giant revealed new agentic capabilities in Search, where users can create, customize, and manage multiple AI agents to stay updated on topics of interest. The announcement is part of Google’s larger push toward agentic AI systems that can take initiative and assist with ongoing tasks instead of answering one question at a time.

Unlike traditional search tools that respond only when prompted, Google’s information agents are designed to operate continuously in the background, 24/7, helping users stay informed about their interests without needing to repeatedly search for the same information every day.

Instead of delivering a list of links, the agents can synthesize information from multiple sources, explain why something matters, compare perspectives, and provide actionable insights. In many ways, the agents represent the next evolution of Google Alerts, the notification service Google launched in 2003. However, these agents are designed to go beyond simple notifications.

For instance, someone following the stock market could create an information agent focused on specific companies, share price, or economic trends. The agent could monitor market activity throughout the day, track breaking news, summarize earnings reports, alert users when major changes happen, and provide summaries and links to learn more.

Image Credits:Google

It could also help with everyday tasks, such as tracking flight prices for upcoming trips, monitoring sports teams and live events, following breaking news, keeping tabs on housing or job market trends, and tracking weather or traffic.

To use the feature, users can open AI Mode in Search and enter a prompt.

For example: “Keep me updated on nearby movie tickets for ‘The Mandalorian and Grogu.’” When something relevant appears, the Google app sends a push notification. You’ll also see your active tracked topics in your AI Mode history, where you can jump back in to manage, refine, or turn off an alert.

Information agents will be available this summer. The company is first rolling them out to Google AI Pro and Ultra subscribers in the U.S., then to additional markets later on.

In addition to these information agents, Google also introduced a major redesign of Search, including what it describes as a reimagined “intelligent search box,” the company’s biggest change to Search in more than 25 years. The new interface is designed to support longer, more conversational queries. There’s also a new AI-powered query suggestion system that goes beyond traditional autocomplete, helping users craft nuanced and context-aware searches.

Catch up on the rest of Google IO 2026’s big news

Google Search as you know it is over

Google updates Gemini app to take on ChatGPT and Claude

Google introduces Gemini Spark, a 24/7 agent assistant with Gmail integration

How to use Google’s new information agents

In Brief

Messaging platform giant Discord has switched on end-to-end encrypted voice and video messaging for every user. The end-to-end encryption feature means that Discord users can now communicate privately without anyone else being able to listen in to their calls, not even Discord.

This is a major privacy win for the community’s hundreds of millions of users, soon after social media giant Meta pulled the plug on Instagram’s end-to-end encrypted messaging feature earlier this year. TikTok said it would also not end-to-end encrypt user messages after it became a U.S. company.

Discord launched the end-to-end encrypted voice and video calling feature in 2024, and on Monday rolled out the feature to every user, no action needed.

“End-to-end Encryption is now standard for every voice and video call on Discord, outside of stage channels. No opt-in required,” said Mark Smith, Discord’s vice president of core technologies, in a blog post.

Three-year-old Mach Industries has acquired solid rocket motor startup Exquadrum in a $50 million cash-and-equity deal, the Huntington Beach-based defense startup tells TechCrunch. Exquadrum — now rebranded as Mach Energetics — has been fully folded into Mach’s operations, giving it direct control over one of the most important, and constrained, components in modern unmanned systems.

The deal itself started with a lucky break. The two companies first connected last September when an Exquadrum customer at an MIT recruiting event happened to overhear a Mach recruiter mention that the company was in the market for a solid rocket motor supplier. Introductions were made, Mach came in as a customer, and now, roughly five months later, it has acquired the company outright, beating out upwards of eight other potential buyers in the process, it says.

“The Exquadrum acquisition marks an important next stage in Mach’s growth,” said founder and CEO Ethan Thornton, who dropped out of MIT at 19 to start the company. “As we deliver vehicles to the warfighter, we’ll continue to vertically integrate our supply chain across solid rocket motors, engines, radar, and avionics to ensure we deliver the best possible product at the lowest cost. In many areas of the defense industrial base, these components are not only too expensive or lacking performance, they’re simply unavailable, with lead times stretching years. In short, vertical integration is non-optional.”

That supply problem is real and getting more acute. Decades of consolidation have left the domestic solid rocket motor market effectively controlled by two large primes — Aerojet Rocketdyne and Northrop Grumman — with little independent capacity to absorb the growing demand that modern drone warfare has created.

Indeed, in February, the Pentagon awarded the defense tech outfit Anduril $43.7 million specifically to expand domestic SRM production (its second such investment in the company in just over a year), explicitly calling SRMs a critical bottleneck in the munitions supply chain.

Mach is now explicitly positioning itself as part of the solution, and not just for its own programs but for the broader ecosystem. Mach Energetics plans to sell components, testing services, and subsystems to other defense firms, a maneuver that suggests Mach sees itself as potential infrastructure for the defense tech industry and not just a systems builder.

According to Mach, all 85 Exquadrum employees are coming over as part of the deal, along with the company’s IP, business lines, and its 70,000-square-foot facility in Victorville, California, which is anchored by a nearby energetics and rocket propulsion test site. The combined company now has roughly 350 employees. Exquadrum co-founders Kevin Mahaffy and Eric Schmidt (no relation to the former Google CEO) are both taking on leadership roles within Mach Energetics and the broader organization.

The acquisition also mirrors moves by other ambitious defense tech startups that are focused on owning the stack and using cost and speed as competitive weapons. Mach has five vehicle programs in various stages of development — Viper, a jet-powered VTOL; Glide, a high-altitude strike glider; Stratos, an airborne surveillance platform; Dart, its low-cost counter-drone interceptor; and Pike, a long-range strike munition built for large-scale deployment — with plans to enter production on at least three this year. The company says the acquisition meaningfully improves unit economics across all of them at exactly the moment it is starting to scale.

Mach has raised nearly $200 million in total — most recently a $100 million Series B last June led by Bedrock Capital, Khosla Ventures, and Sequoia Capital — at a valuation of $470 million. It’s a multiple that right now looks modest for a company with this trajectory and will definitely be worth watching as the rubber starts to meet the road this year.

Shay Shwartz knows a lot about email phishing attacks. As a teenager, he made money as a hacker, but after getting caught at age 16, he realized he could use his cyber talents to prevent attacks rather than launch them.

He went on to spend about a decade in top-tier cybersecurity roles, leading major projects for Israel’s elite defense and intelligence units, including work connected to the Iron Dome project, before joining Axis, the startup later acquired by HPE.

All along, he had been itching to launch his own startup, and two years ago, he finally took the plunge.

His startup Ocean, an agentic email security platform built to fight AI-powered attacks, just emerged from stealth mode with $28 million in total funding. The round was led by Lightspeed Venture Partners, with participation from Picture Capital and Cerca Partners. High-profile angel investors also joined the round, including Wiz co-founder and CEO Assaf Rappaport, as well as Yevgeny Dibrov and Nadir Izrael, the co-founders of Armis, which recently sold to ServiceNow for $7.75 billion.

While established vendors like Proofpoint and Mimecast, along with newer players like Abnormal Security, help detect standard phishing attacks, Shwartz (pictured right next to co-founder and CTO Oran Moyal) argues that AI requires a different defensive approach.

In the past, only highly sophisticated hackers could pull off spear-phishing due to the sheer amount of time, research, and manual labor needed to launch targeted attacks.

“AI just made the entire process automatic, so the scale is much, much bigger now,” Shwartz told TechCrunch. “I can instruct LLM to go and understand exactly who you are, harvest large amount of public information, and create those phishing attacks very targeted against you.”

Ocean claims its AI can thoroughly analyze the context of every incoming email to detect fraud and impersonation attempts.

The startup is already reviewing billions of emails each month for customers including Kayak, Kingston Technology, and Headspace.

Shwartz said Ocean built a small language model tailored to quickly analyze emails, understand the sender’s intent, and evaluate it against the user’s specific organizational context.

“This is like having a guard in every door,” Shwartz said. “This is how we make the inbox a safe place with high hygiene.”

The jury’s speedy decision to reject Elon Musk’s lawsuit against the other founders of OpenAI and Microsoft confirmed what we saw in the courtroom: Musk’s case was a weak one, in part because he waited so long to file it.

Watching the closing arguments last week, OpenAI’s attorneys detailed point-by-point how the law was on their client’s side, while the plaintiffs team focused on Sam Altman’s apparent lack of credibility and expressed disbelief that anyone would disagree with Musk’s accusations.

The final effect was that, after the verdict, some found it hard to believe Musk had lost — including the man himself. In a post he later deleted, Musk called Judge Yvonne Gonzalez Rogers a “terrible activist Oakland judge,” then announced his plans to appeal, declaring “there is no question to anyone following the case in detail that Altman & Brockman did in fact enrich themselves by stealing a charity.”

But Altman and Brockman weren’t the only figures who benefitted from OpenAI’s non-profit investments. As much as Musk and his legal team tried to make the trial about Altman, the proceedings revealed just as much about Musk himself.

One incident that came out in court showed Musk benefiting from OpenAI in an uncomfortably familiar way. Greg Brockman testified that in 2017, Musk asked him to bring a team of OpenAI researchers down to Tesla’s headquarters to help with the autopilot team for a few weeks. “It was pretty clear that was not something we could say no to,” Brockman said.

Brockman described taking a team of leading scientists, including Andrej Karpathy, Ilya Sutskever, and Scott Grey, to consult with the “demoralized” Tesla workers. They helped come up with ideas to improve the vehicle’s self-driving technology, with Sutskever telling the team that if they could find 10,000 images of a tricky corner case, they would be able to fix their software. Musk even asked Brockman to recommend employees to fire, which he declined to do.

Another person familiar with the episode confirmed Brockman’s account, and said Tesla did not reimburse OpenAI for the time and effort of its employees. Musk’s family office, Excession, didn’t reply to a request for comment.

The heart of Musk’s case is that Altman, Brockman and OpenAI committed a “breach of charitable trust” — that Musk donated funds for a specific charitable purpose, and his cofounders instead used them for something else. He also accuses them of “unjust enrichment” due stock and other benefits from OpenAI’s for-profit.

In the case of the OpenAI scientists parachuting into Tesla, Musk’s charitable donations were intended to hire scientists focused on securing the benefits of AGI. Instead, he had them work for free at his for-profit company.

Dorothy Lund, a Columbia Law School professor and the co-host of the Beyond Unprecedented podcast, told TechCrunch that this arrangement wouldn’t be legal, calling it “a bit rich for Musk to be suing for breach of a charitable trust, when he appears to have been redirecting assets in a way that was inconsistent with that mission.”

It’s true that the self-driving work involved artificial intelligence, but witnesses for Musk emphasized that Tesla’s self-driving project was very different from OpenAI’s research agenda. That’s in part because Karpathy left OpenAI for Tesla shortly after this incident. OpenAI’s attorneys portrayed the departure as Musk violating his duty to the lab, where he was co-chair of the board, by recruiting one of its key researchers to his own company.

The other fact that no doubt influenced the jury was the amount of time Musk spent trying to gain sole control of a potential OpenAI for-profit affiliate in 2017. Musk deployed good cop, bad cop tactics in an attempt to convince his cofounders to let him have total control of OpenAI’s for-profit affiliate — giving them free Teslas, and threatening to withhold his donations.

His efforts put his attorneys in a tricky spot, facing a need to convince the jury there was a significant difference between what Musk envisioned, and the for-profit that was ultimately created. They suggested a “small adjunct” for-profit would be permissible, though OpenAI’s witnesses showed non-profits with large commercial arms are common.

Indeed, there’s a very plausible counter-factual where Musk took one of the offers his cofounders made to split their equity more evenly, and finds himself today as one of OpenAI’s largest shareholders — just not the controlling one. But several times during the trial, Musk’s associates testified that he refuses to invest in any business he could have sole control over.

The failure of Musk’s claims because he filed them too late has been cited as a technicality, but the statute of limitations has substance behind it: People and businesses make important decisions and spend resources based on their understanding that what they are doing is permissible. If someone like Musk waits too long to sue, then the cost of unravelling all those decisions can outweigh a just reimbursement.

No members of the jury have spoken about how they arrived at their verdict. However, they were asked to consider if, before Aug. 5, 2021, Musk should have known that OpenAI was spending resources outside its mission or launching for-profit affiliate. The answer to that is clear: Musk himself was doing those things.

Google is getting (back) into the smart glasses game.

At Google I/O on Tuesday, the company announced a new partnership with Warby Parker and Gentle Monster to produce a new line of AI-powered glasses. The company says that the devices will be built to pair with Android and iOS devices and were designed in collaboration with Samsung. They will be available later this year, the company said.

Google is calling the new devices “audio glasses,” in that users will be able to issue verbal commands to them and get things done via its ecosystem of apps and services, including Gemini.

Image Credits:TechCrunch

The user simply talks to their glasses (the demo shared on Tuesday involved a Googler ordering a coffee online by merely talking to the glasses), and the device, when synced, complies.

Google has dabbled in smart glasses a number of times over the years. It notoriously launched Google Glass years ago, which ultimately helped spawn the derogatory term “glassholes.”

The smart glasses space has changed a little bit since then, however. Lately, major companies — most notably Meta — and a small army of startups and smaller firms, have invested in the space.

Catch up on the rest of Google IO 2026’s big news

Google Search as you know it is over

Google updates Gemini app to take on ChatGPT and Claude

Google introduces Gemini Spark, a 24/7 agent assistant with Gmail integration

How to use Google’s new information agents

We’ve all pulled up Street View on Google Maps to show a friend what our childhood home looked like, or dropped that little person icon onto the streets of Paris to see if we booked a hotel in a cool neighborhood. Imagine being able to do that, but in a more immersive, interactive way that allows you to really simulate the street and its environs, and even do things like adjust the weather or see what it would look like in a “Day After Tomorrow” scenario.

That’s one of the goals of Google’s latest integration. Starting today, Google DeepMind is connecting Street View to Project Genie, the company’s general-purpose world model that can generate diverse, interactive environments. The new feature launched during the Google I/O 2026 developer conference.

“It’s really powerful for both the agent [and robotics] use case and for humans to play with, and that’s always been the thesis of Genie,” Jack Parker-Holder, a research scientist on DeepMind’s open-endedness team, told TechCrunch.

He gave the example of a new robot being deployed in London, which rarely sees the sun. Genie could, Parker-Holder says, simulate those scarce occasions when the sun glints off the Victorian housing, so the rays don’t shock the robot when it happens.

Loading the player…

“Simultaneously, you might say, ‘I’m going to New York City, but not this time of year,’” he continued. “‘It’s going to be snowy. I want to see what that block looks like in the snow.’”

Google has been collecting Street View data for 20 years via cars with cameras and individuals strapped with “tracker backpacks.” The tech giant has collected north of 280 billion images across 110 countries and seven continents.

“With Street View, we have imagery from a large quantity of the world,” Jack said. “You can imagine how potentially powerful it is to combine this rich source of real-world information and data with an ability to simulate worlds.”

Google released its latest world model Genie 3 for research preview last August and opened up access to the tool to Google AI Ultra subscribers in the U.S. in January, allowing customers to create interactive game worlds from text prompts or images. The goal is to use Genie for educational experiences, gaming, and robotics training.

Genie 3 is already helping to power one of Waymo’s simulators to train its self-driving cars on “exceedingly rare events” like tornadoes or casual elephant encounters. Adding Street View data to that could help Waymo prepare to launch in more cities around the globe.

Waymo has its own simulator that it relied on to scale to 11 U.S. cities and test its AI driver in several more. The difference with Genie, says Parker-Holder, is that those are all from the car’s point of view. Street View allows for not only simulating a world anchored to a real place, but also shifting the point of view to other types of agents, like a human or a robot.

Google is launching Street View in Genie to some Ultra users in the United States starting today, with access rolling out at scale over time. Global Ultra users will gain access over the next few weeks, per the company.

The researchers’ goal is to put this new capability into as many hands as possible, per Diego Rivas, a product manager at DeepMind. He cautioned that Street View in particular and Genie in general is still an experiment, so there’s much to improve upon in terms of accuracy.

In the samples the Google team showed me — including an underwater simulation of a neighborhood I used to live in — the results are impressive and recognizable, but still video game quality rather than photorealistic. The models are also not yet physics-aware, meaning they don’t yet understand cause and effect. For example, in a simulation of a woman running through a snowy Joshua Tree, she ran right through cacti and bushes.

Compare that to, say, Google’s image generator Nano Banana — which can now generate perfect text in infographics — or its video generator Veo — which understands that paper boats drift on water currents, smoke disperses into the air, and fabric drapes over forms.

Physics isn’t hard-coded into these models; they learn it intuitively over time through passive observation, as a living being would.

“I think for this kind of model, it’s maybe six to 12 months behind video in terms of the accuracy and quality, so I think it’s something we will solve,” Parker-Holder said.

Jonathan Herbert, director of Google Maps who started on the Street View team as an intern 12 years ago, said that Genie can’t yet create a faithful reconstruction of a street. He thinks the real breakthrough is the AI’s spatial continuity. If you turn 360 degrees, the AI correctly remembers and simulates the environment behind you. From that point on, the model can build a new environment on top of that.

“We have long thought about how we can build out the best and richest model of the world on top of Street View data,” Herbert said. “It’s definitely been an idea of ours to use Maps Data in new ways and for new kinds of AI research for a pretty long time.”

Catch up on the rest of Google IO 2026’s big news

Google Search as you know it is over

Google updates Gemini app to take on ChatGPT and Claude

Google introduces Gemini Spark, a 24/7 agent assistant with Gmail integration

How to use Google’s new information agents

Google launched on Tuesday Gemini 3.5 Flash, a new AI model that the company says is its strongest yet for coding and autonomous AI agents. The model, which was introduced at the company’s annual Google I/O developer conference, can independently execute coding pipelines, manage research projects, and, in internal tests, build an operating system entirely from scratch.

The release signals Google’s shift from pitching AI as a conversational tool to AI as an agentic tool. It’s not just answering questions, but planning, building, and iterating on real work with minimal human input.

Koray Kavukcuoglu, DeepMind’s chief technologist, told reporters on Monday ahead of the public launch: “3.5 Flash offers an incredible combination of quality and low latency. It outperforms our latest frontier model, 3.1 Pro, on nearly all the benchmarks,” including coding, agentic tasks, and multimodal reasoning.

He added that it is 4x faster than other frontier models, a speed that’s ideal for coding and agentic tasks, but that Google has “taken it to another level” by developing an optimized version of Flash that’s 12x faster with the same quality.

That speed is central to Flash’s design for agentic work, where multiple AI agents run at the same time on long-running tasks, according to Kavukcuoglu. Onstage at I/O, Google engineer Varun Mohan, demonstrated agents spawning off to work on separate components before coming together to build a full operating system inside Antigravity, the company’s agentic development platform and IDE.

Image Credits:Google

Kavukcuoglu said Flash 3.5 was co-developed with Antigravity so that agents could have a “native environment where they can live, work, and execute.”

At I/O, Google released Antigravity 2.0, a stand-alone desktop application designed around agent-first development.

The gains are showing up beyond demos. Google says 3.5 Flash’s agentic capabilities are already creating impact among partners, like banks and fintechs automating multi-week workflows, or data science teams finding insights in complex data environments.

The model can run autonomously for multiple hours, though Tulsee Doshi, Google’s senior director and head of product, said it will at times pause and ask for user input when it hits a decision point or permission issue that requires human judgment.

When Google releases its forthcoming 3.5 Pro model, the two are designed to work in tandem.

Doshi told TechCrunch that “3.5 Pro becomes your orchestrator, your planner, and then it actually can leverage Flash to be the various sub-agents. I think it really comes down to where do you really want that reasoning power, where you actually want that larger model that can really push on the reasoning side versus where do you have tasks that really do merit good brute force tool use capabilities?”

Now, 3.5 Flash is the default model in the Gemini app and in AI Mode in Search globally. At I/O, Google also announced agentic capabilities coming to Search, letting users create, customize, and manage AI agents directly on the platform. The new model will also power Gemini Spark, Google’s new personal AI agent designed to run 24/7 to help consumers manage their digital life.

Providing that level of AI capability for average consumers comes with scrutiny. Google is currently facing a lawsuit after a man nearly committed a mass casualty event and died by suicide following weeks of chatting with Gemini last year.

The implications for harm only grow when making powerful autonomous agents available more broadly. Google says Gemini 3.5 has strengthened cyber and CBRN (chemical, biological, radiological, and nuclear) safeguards and is better calibrated to engage with sensitive questions rather than refuse them outright.

Gemini 3.5 Flash is available generally today via Antigravity, the Gemini API, and Gemini Enterprise, as well as through the Gemini app and AI mode in Search.

The era of the “ten blue links” is officially over.

At its Google I/O conference on Tuesday, Google unveiled an AI-powered overhaul of Search centered around a reimagined “intelligent search box” — what the company describes as the biggest change to this entry point to the web since the search box debuted more than 25 years ago.

Instead of returning a simple list of links, Google Search will drop users into AI-powered interactive experiences at times. Google is also introducing tools that can dispatch “information agents” to gather information on a user’s behalf, along with tools that let users build personalized mini apps tailored to their needs.

Loading the player…

The resulting experience will no longer look much like how people envision Google Search, which has long been defined by ranked links to websites that have the information you need.

With the revamped Search experience, the new search box simply expands to accommodate longer, more conversational queries, rather than making you decide what type of search experience or mode you want to choose at the start of your query. It will also have a new AI-powered query suggestion system that goes beyond autocomplete to help people craft more complex and nuanced queries, Google says.

Google’s AI Overviews will also allow users to ask follow-up questions in AI Mode, beginning Tuesday, the company noted.

Image Credits:Google

Google is also introducing agentic capabilities and AI-powered interactive features into the search experience. This means people will spend even less time clicking the traditional blue links that Google Search used to return.

Starting this summer, people will be able to create, customize, and manage multiple new “information agents” within Google Search. These agents can work in the background 24/7 to track changes on the web and alert you to new information. For instance, you could have an agent track market movements based on customer parameters, Google suggests.

Image Credits:Google

While the underlying technology here is powered by AI, which makes it more capable, the idea itself is not a new one.

In 2003, Google launched Google Alerts, a change-detection service that emailed users when new web results matched their search terms. The web was smaller and more manageable then, of course, so this became a part of many information workers’ tool sets. (That service still exists in some form but is no longer the way most web users go about acquiring new information.)

Information-gathering agents are an evolution of Google Alerts. Beyond spotting changes, they can make sense of them, too.

“You could send an alert to track market movements in a particular sector with very specific parameters, and the agent will map out a monitoring plan for you, including the tools and the data it needs to access — like our real-time finance data,” Google’s head of Search, Liz Reid, explained in a press briefing. “And it will then keep track of those changes and let you know when the conditions are met, and provide a synthesized update with links and information you can dive into further,” she added.

Image Credits:Google

This shift means that “searching the web” will increasingly be performed by AI agents rather than humans. Instead, people will focus more on acting on the information those agents provide instead of manually clicking links.

Image Credits:Google

Links will become an afterthought with the coming changes to the Search results experience, which builds on Google’s earlier launches of AI search features, like its short summaries known as AI Overviews and its conversational search, AI Mode.

AI Overviews are now used by more than 2.5 billion monthly users; meanwhile, its conversational search mode, launched last year, now tops 1 billion monthly users. (ChatGPT, for comparison, has 900 million weekly active users, as of earlier this year. This suggests that ChatGPT is now seeing more frequent engagement, with users coming back repeatedly throughout the week, while Google has more total unique people touching its AI features over the course of a month.)

Now, thanks to a combination of Gemini and Google Antigravity, the company’s agentic development platform, Search results will begin to look more like interactive web pages.

Image Credits:Google

“Search can build custom experiences just for your individual questions, from dynamic layouts, interactive visuals to persistent and stateful project spaces that you can return to again and again,” says Reid. One of the ways Google is integrating these new capabilities is with “generative UI” (user interface), where it builds custom widgets and visualizations on the fly in answer to users’ search questions.

You can imagine, for example, how a question about black holes in space could lead to an interactive visual that brings the concept to life, Reid said, adding that users can then ask follow-up questions and see Google respond with brand-new visuals in real time.

Image Credits:Google

Google says the new system was built in partnership with the Google DeepMind team and uses Gemini Flash 3.5. It will roll out to everyone who uses Google, free of charge, this summer.

In addition, Google will allow users to tap into Antigravity to build their own customizable, stateful experiences — think “mini apps” — directly in Search using natural-language commands. Again, this isn’t so much about information retrieval as it is about action. For instance, you could build a meal-planning app using information from your own calendar to help you decide what to prep and when to eat, or a fitness app created for your specific goals.

Image Credits:Google

Combined, these changes will likely further decimate Google referrals to publishers, which have already been suffering from declining referrals due to AI Overviews. This has put some ad-dependent media operations out of business, and now things will likely get worse.

There’s little time left for publishers to adapt. The new search box is arriving this week, and generative UI is arriving this summer. Both are free. The mini-app-building feature and information agents will roll out first to Google AI Pro and Ultra subscribers this summer.

But Google’s long-term plan is to make its AI technology more broadly accessible, including its personal AI agent Spark, which will eventually be free, as will many of the AI features.

“Part of the reason we focus on delivering frontier models — highly capable, but also very efficient, fast, and at a lower price — is because we want to bring it to as many people as possible, and so I think that’s an area where we will shine,” Google CEO Sundar Pichai said in a press briefing ahead of I/O.

The AI coding boom is now coming directly for Android app development. On Tuesday at Google IO 2026, the company announced new native Android app creation capabilities in its web-based Google AI Studio, shrinking a process that takes weeks of setup and coding down to minutes.

The company also said that consumers will be able to use Gemini AI to find the apps they need, both on the Play Store and the web, expanding opportunities for developers to have their apps discovered.

Google says the new capabilities could make sense for anyone from a seasoned developer looking to prototype a new app quickly to a first-time creator.

By offering the ability to essentially vibe-code Android apps via web-based tools, Google is ramping up the competition with other AI-powered development tools, like Cursor, Replit, Lovable, Claude Code, and others, while also opening up Android development to a new type of user: a non-technical creator. The news also represents an expansion of Google’s earlier addition of AI-powered coding with Gemini in its desktop version of Android Studio.

The apps are built with the Kotlin programming language using Google’s Jetpack Compose toolkit and with support integration with hardware sensors like GPS, Bluetooth, and NFC, the company says. However, the resulting creations, for now, are only meant to be used personally, as publishing for family and friends is still on the roadmap.

The company suggests the technology could be used for the creation of personal utilities and simple social apps, hardware-enabled experiences, or AI-powered experiences.

For now, would-be app developers can use the embedded Android Emulator directly in a web browser to preview and interact with the app as it’s being built. Users can then install the app on their Android phone over a USB cable connected to their computer, using the integrated Android Debug Bridge (adb).

For those looking to take their project further, AI Studio can automatically create the app record, package the bundle, and upload it to an internal testing track in Google Play Console for developers. This allows users to continue to iterate on their app while updating on their devices along the way.

Those who want to take the next steps to publishing the app more publicly can hand off this version of the project to Android Studio by downloading a zip file and exporting it directly to GitHub. In time, Google plans to allow creators to publish their apps for use by family and friends and will add support for Firebase integrations (Firestore, Firebase Auth, Firebase App Check, and other tooling).

In doing so, the company is imagining an Android app ecosystem where users find apps from among their own network of friends, not just the Play Store.

However, for the latter, Google is infusing AI into the experience here, too.

A new “Ask Play” AI-powered overlay will allow users to discover new apps by having natural conversations with AI within the Play Store.

Image Credits:Google (screenshot)

Perhaps more importantly, apps will begin to be surfaced with users’ conversations with Google’s Gemini virtual assistant, exposing developers’ apps to millions of users. This will roll out in the weeks ahead across Gemini on the web and on Android. Later this year, Gemini will also surface over 450,000 movies and TV shows, plus where to livestream sports, which can directly link users from their queries to a developer’s Android app with the content in question.

Image Credits:Google

While Google previewed a number of Android-related announcements last week, it held back on sharing the news of the native Android app development until Tuesday’s start of its annual developer conference, Google I/O. That suggests the company believes this is bigger news and more closely tied to its idea of putting AI to real-world use, as was the larger theme of this year’s event, where AI was spread across Google products, from workspace productivity apps to AI tools, search, mobile apps, and more.

Catch up on the rest of Google IO 2026’s big news

Google Search as you know it is over

Google updates Gemini app to take on ChatGPT and Claude

Google introduces Gemini Spark, a 24/7 agent assistant with Gmail integration

How to use Google’s new information agents

At its annual Google I/O event on Tuesday, Google announced that the Gemini app is getting a series of new updates, including a “Daily Brief” feature, a redesigned interface, access to a new AI video model called Gemini Omni, and new personal AI agent called Gemini Spark.

The updates signal Google’s push to turn its Gemini app into an all-purpose AI hub rather than a stand-alone chatbot and to make the AI assistant more competitive with apps like ChatGPT and Claude.

Google describes the new Daily Brief feature as a personalized digest designed to be your first stop each morning. It pulls together information from a user’s inbox, calendar, and most important tasks, and then organizes them into a clear overview. The company says Daily Brief doesn’t just summarize this information — it also prioritizes tasks and suggests next steps, with the most important items shown first.

Daily Brief is rolling out today to Google AI subscribers in the United States.

Image Credits:Google

The Gemini app already has considerable reach: The company said it has more than 900 million monthly users and is available in over 230 countries and in more than 70 languages. But Google clearly wants more and without losing its existing customers.

Google said it rebuilt the app from the ground up. Now when users open the app, they’ll see a new design language called “Neural Expressive,” which brings fluid animations, vibrant colors, new typography, and haptic feedback.

Image Credits:Sarah Perez/TechCrunch

Gemini’s responses are no longer presented as a wall of text, which is the case with most AI chatbots. Instead, key information appears in bold at the top. Additional text and possibly other elements such as images and timelines appear as the user scrolls down.

As for Gemini Spark, Google describes it as a 24/7 personal AI agent that helps you navigate your digital life. Spark transforms Gemini from an assistant into an active partner that does real work on your behalf. Since it’s a cloud-based agent, Spark keeps working in the background even when you lock your phone. With Gemini Spark in the Gemini app, users will be able to create their own custom workflows.

Spark is currently in testing and the company expects to make it available to Google AI Ultra subscribers next week.

Image Credits:Sarah Perez/TechCrunch

The company’s new AI video model, Gemini Omni, combines Gemini with Google’s generative media models to create outputs grounded in knowledge. For example, you could give it a simple prompt like “claymation explainer of protein folding.” The model lets you upload audio, images, and video to generate a consistent, high-quality video, according to Google.

By adding access to a new video-generation model like Gemini Omni, Google is intensifying competition among major AI platforms in the ongoing race to lead multimodal content generation. The model is rolling out to Google Flow and YouTube Shorts for Google AI subscribers, underscoring the company’s broader push into multimodal content creation and AI-powered video tools.

Catch up on the rest of Google IO 2026’s big news

Google Search as you know it is over

Google updates Gemini app to take on ChatGPT and Claude

Google introduces Gemini Spark, a 24/7 agent assistant with Gmail integration

How to use Google’s new information agents

In Brief

In addition to serving up its own flavor of vibe-coding tools for building on Android, Google announced it will also offer tools that allow AI agents, like Claude Code, OpenAI’s Codex, or its own Antigravity or Gemini in Android Studio, to accelerate Android app development.

Google shared on Tuesday at its annual Google I/O developer conference that its Android CLI (command line interface) is now stable at version 1.0 and can be used by those who are leveraging AI agents to build for them — no matter what their preferred coding platform may be.

The move acknowledges that many people are now building for Android with AI agents that aren’t from Google. The company is trying to find a way to make its specialized knowledge, like what’s found within Android Studio, more accessible.

With Android CLI, AI agents can retrieve knowledge about Android development via a new “android studio” command that can tap into the capabilities of Android Studio when building their own project. From there, the agents will be able to tap into a range of other commands and tools.

Google Antigravity, the company’s agentic development platform, will include an optional bundle that will install the tools and knowledge included in Android CLI, so it can also perform the core tasks for Android app development, Google says.

At Google I/O on Tuesday, Google introduced Universal Cart, its agentic hub for managing shopping in one place. The tech giant also announced updates to its Agent Payments Protocol (AP2) and teased that it would bring the technology to Google products in the coming months, enabling users to authorize agents to make payments on their behalf.

The announcements signal Google’s push to turn AI assistants from passive recommendation tools into active participants in online commerce. By launching a centralized shopping system and building infrastructure that lets software agents complete purchases autonomously, the company is positioning itself to control more of the entire shopping journey, and potentially the relationship between consumers and the merchants competing for their attention.

With Universal Cart, users can add products they’re considering from anywhere on Google — while browsing Search, chatting with Gemini, watching YouTube, or reading Gmail. Once items are added, Universal Cart tracks deals, monitors price drops, surfaces price history insights, and alerts users when items are back in stock.

Image Credits:Sarah Perez/TechCrunch

The feature is built around something Google knows well, which is that most people shop across multiple devices, multiple retailers, and over the course of many days.

The cart also uses AI to help shoppers make better decisions. For example, if you’re building your first custom PC, you can add parts from multiple merchants into a single cart, and Google may flag compatibility issues, such as a processor that doesn’t work with the motherboard you selected — and suggest an alternative.

For frequent travelers or rewards maximizers, the feature can also surface hidden savings and help stretch your points further because it’s built on Google Wallet.

Image Credits:Google

Thanks to Google’s open-standard Universal Commerce Protocol (UCP), users can check out directly through Google with participating merchants, or transfer their items to the merchant site and complete the purchase there.

Universal Cart is rolling out in the U.S. today and coming to the Gemini app this summer, with YouTube and Gmail to follow, Google says.

Google also announced that UCP is expanding to more categories, like hotels and local food delivery services. UCP-powered experiences will also expand beyond the U.S. to Canada and Australia in the coming months, and later to the U.K.

The more consequential announcement for the commerce industry may be AP2, Google’s protocol designed to let AI agents securely make payments on users’ behalf within defined limits. At I/O, Google detailed the guardrails users can set, including specifying the brands and products they want, and a spending limit. When those conditions are met, the agent makes the purchase automatically.

Image Credits:Sarah Perez/TechCrunch

Google says it’s bringing AP2 to its own products in the coming months. That integration would give Google direct visibility into what consumers discover, consider, and ultimately buy, and it’s a degree of commercial influence that retailers and payment processors will be watching closely.

Under the hood, AP2 creates a transparent, verifiable link between the user, the merchant, and the payment processor, with encryption protecting user data throughout. The protocol also includes tamper-proof digital records that ensure the agent is always acting on the user’s behalf, and a permanent audit trail that both buyers and sellers can reference for returns or disputes.

Catch up on the rest of Google IO 2026’s big news

Google Search as you know it is over

Google updates Gemini app to take on ChatGPT and Claude

Google introduces Gemini Spark, a 24/7 agent assistant with Gmail integration

How to use Google’s new information agents

In the race to build compelling personal AI agents, Google may have an underrated advantage: It already has all your emails.

At its Google I/O developer conference on Tuesday, the company announced a new agentic personal assistant called Gemini Spark that was built from Gemini base models and an agentic harness from Google Antigravity.

Alphabet CEO Sundar Pichai described Spark as the next evolution of smart digital assistants, using agentic AI to take on long-horizon tasks with minimal oversight.

“It’s your personal AI agent that helps you navigate your digital life, taking action on your behalf and under your direction,” Pichai told reporters during a pre-briefing of the product. “It runs on dedicated virtual machines on Google Cloud seamlessly, [so] you don’t need to keep your laptop open to make sure it’s running.”

Spark follows a wave of popular agentic products from major AI labs, most notably Anthropic’s Claude Cowork and OpenAI’s ChatGPT agent, but it will have particular value because of its integration with the larger Google suite of products. Spark will include out-of-the-box integrations with Gmail, Google Docs, and other Google Workspace products, saving users the work of setting up connections and permissions with outside apps.

Users can email Spark directly through a dedicated Gmail address, and the agent can interact with the web directly through Chrome. On mobile, you’ll also be able to track the agent’s progress through the new Android Halo system.

“Need to send an email to your boss with a status update? Spark can pull all the facts from your emails, your docs, your sheets, and slides and write the draft for you,” said Google Labs’ Josh Woodward, VP of the Gemini App and AI Studio. “Small businesses are using Spark. They can watch over their inbox, so they never miss a question from a customer.”

Like other agentic assistants, Spark can be integrated into a wide range of services over MCP, and Google expects to roll out more connections in the months to come.

Spark is currently in testing at Google; the company expects to make it available to Google AI Ultra subscribers next week.

Catch up on the rest of Google IO 2026’s big news

Google Search as you know it is over

Google updates Gemini app to take on ChatGPT and Claude

Google introduces Gemini Spark, a 24/7 agent assistant with Gmail integration

How to use Google’s new information agents

At the Google I/O developer conference, Google announced it’s bringing a voice-based prompting feature to Workspace apps such as Docs, Keep, and Gmail. These features can help you create drafts, take notes, and search for emails.

In Docs, you can create a draft document using your voice. For instance, in a demo, Google showed that a user can fetch résumé details from Drive, add event logistics from an email on top of the document, and even include some humorous anecdotes.

Image Credits:Google

Previously, users would have to type all this out, and it’s possible that, while typing, they might write short sentences and then various follow-ups, resulting in a multi-turn conversation that took time. Google’s idea is that with voice, you can use long sentences or ask for multiple tasks in one go. Plus, the feature understands when you change your mind and ask for a detail to be changed in the same conversation turn.

Google CEO Sundar Pichai said that in the future, users will be able to create and edit documents using their voice.

Loading the player…

What’s more, Google is adding a way to use your voice to dump your thoughts into Keep, and the app will use AI to turn that transcription into a structured note or a list. Notetaking apps like Voicenotes and AudioPen added this kind of feature a few years ago. Lately, dictation apps such as Wispr Flow, Monologue, and Aqua have built this functionality into their voice-based typing products.

Earlier this month, Google released its own dictation product called Rambler, which is built into Gboard and works across apps.

Besides Docs and Keep, the company is adding voice-based functionality to Gmail. With the new feature, users will be able to converse with Gemini and ask for details like their next flight, the code to their Airbnb booking, or time for their physician appointment.

Tech companies are cramming AI into all products and features, and as a result, users are getting attuned to asking lengthier queries.

In some cases, voice is an easier input method to blurt out long sentences and queries describing complex multistep requests. What’s more, the current generation of models is good enough to understand if users change their mind about something in the middle of a sentence and output the final query accordingly. Google is paying attention to this trend and adding more voice-based features across its apps.

Catch up on the rest of Google IO 2026’s big news

Google Search as you know it is over

Google updates Gemini app to take on ChatGPT and Claude

Google introduces Gemini Spark, a 24/7 agent assistant with Gmail integration

How to use Google’s new information agents

At Google IO 2026, the company unveiled a new version of its agentic coding app, Google Antigravity 2.0, with an updated desktop app, a CLI tool, and an SDK for custom workflows. The company launched its Antigravity tool last year as a response to agentic coding software such as Cursor.

The company said that with the new desktop app, users can orchestrate multiple agents and execute tasks simultaneously. Plus, you can design custom subagent workflows and schedule tasks that can automatically run in the background. The new app also easily lets you integrate projects with Google AI Studio, Android, and Firebase.

A lot of this is powered by the company’s new Gemini 3.5 Flash model, which was co-developed using Antigravity, Google said.

Google is also adding native voice command support to Antigravity, just like it has added to multiple consumer products, including Gmail and Docs. And it’s launching a new Antigravity CLI (command line interface) tool for programmers who want to use a terminal for creating agents. Google is asking users who used the Gemini CLI tool to migrate to the new tool.

What’s more, Google is launching an Antigravity SDK for developers to build custom agents based on Google’s coding tool. The company will allow Google Cloud customers to connect to Antigravity to build projects. It noted that it will release custom agent templates in AI Studio for enterprise users to get started.

Google is also adding an Antigravity export tool to AI Studio for developers to export their existing project and carry on the work locally.

Google is using Antigravity’s coding chops in consumer products like Search, where users will get a custom UI generated in real time as part of an answer. Here, users will be able to build mini-apps while exploring a topic within search, the company said.

The company is introducing a new AI Ultra plan priced at $100, offering 5x higher AI limits in Antigravity than the Pro plan.

Image Credits:Google

It is also reducing the price of its top AI Ultra plan from $250 to $200, which allows for 20x higher limits than the Pro plan. Other AI labs, including Anthropic and OpenAI, have added both $100- and $200-per-month plans in the last few years to create a tiered system for users with different AI usage requirements.

Catch up on the rest of Google IO 2026’s big news

Google Search as you know it is over

Google updates Gemini app to take on ChatGPT and Claude

Google introduces Gemini Spark, a 24/7 agent assistant with Gmail integration

How to use Google’s new information agents

When Google launched Gemini three years ago, the goal was to build a multimodal large language model — a single neural network that was trained on text, image, audio, and video and could generate content in any of those formats.

Today, at its Google I/O developer conference, the company took a concrete step toward that goal with Gemini Omni, a new family of multimodal models that Google CEO Sundar Pichai says will be able to “create anything from any input.”

Omni will start with video. Users can now combine images, audio, video, and text, and rather than simply stitching those inputs together, Omni reasons across all of them to produce a consistent output. The result is high-quality videos that reflect an understanding of physics, culture, history, and science.

Omni also lets users edit photos with plain text commands rather than complex editing software, similar to Google’s Nano Banana.

Google already has a dedicated video model, Veo, that lets users turn text and images into videos, and even direct and customize avatars. But Google DeepMind director of product management Nicole Brichtova says that today’s release is more than a Veo update: “It’s the next step towards the progression of combining the intelligence of Gemini with the rendering capabilities of our media models.”

One example that Koray Kavukcuoglu, DeepMind’s chief technologist, gave reporters during a media briefing on Monday: When Omni was given a simple prompt like “a claymation explainer of protein folding,” it quickly rendered a video of a stop-motion explainer with a voice-over that said, “Proteins start as chains of amino acids. They fold into patterns like the alpha helix and flat sections called beta sheets, forming a perfect three-dimensional shape.”

The long-term vision for Omni is broader, involving the model being used to do things like generate images from audio, or audio from video.

“When we first announced Gemini, it was our first AI model to be natively multimodal,” Pichai said during the briefing. “We knew that training it on a combination of text, code, audio, images, and video would give it a deeper understanding of the world. With world models, AI is moving from predicting text to simulating reality. Gemini Omni is the next step in that direction.”

As part of the release, users will also be able to create videos with their own digital avatars — something OpenAI popularized on its now-defunct Sora app with Cameos. To prevent deepfakes, users will have to go through a dedicated product onboarding, which involves recording themselves and speaking out a series of numbers, per Brichtova. The avatar then gets stored for future use.

Additionally, all videos created with Omni will include Google’s SynthID digital watermark, which allows users to verify if videos were generated via the Gemini products.

The first model in the family is Gemini Omni Flash, which will roll out today to the Gemini app, YouTube Shorts, and AI creative studio Flow. Flash will be capable of rendering 10 seconds of video, which Brichtova says isn’t a model limitation, but rather a decision based both on a desire to get it into more hands and an anticipation that most users won’t want to make much longer videos yet. Longer video durations are in the pipeline for the near future, though.

Google seems to be pitching Omni Flash as more of a consumer tool. The examples Brichtova and Gabe Barth-Maron, a research engineer at DeepMind, gave on a call with TechCrunch of uses for digital avatars were all personal: Making a video of yourself winning an award or going to the moon, or removing a passerby from the background of a video you took on vacation.

Barth-Maron put it more simply: “They’re like personalized memes.”

“We definitely did focus on making this easy to use for consumers,” Brichtova said. “Not many video models have breached that chasm with consumers, so this is our play to do that.”

The ease of use comes with a caveat: Brichtova and Barth-Maron noted that editing prompts will need to be highly specific, otherwise Omni risks over-editing or unintentionally altering elements the user wanted to keep — a problem Nano Banana users would have run into.

Image Credits:Google

Despite the near-term consumer focus, Omni’s enterprise and creative implications are obvious, and Google will make Omni available via API in the coming weeks. The avatar-generating tool — a capability that is available today on Shorts — is something Google expects content creators to pick up. But more broadly, an end-to-end multimodal workflow could be transformative for advertisers and filmmakers.

Startup Luma AI is building something similar, an agentic tool that can generate an entire ad campaign based on a short brief and a product image, powered by its own “unified” model.

“We’re actually pretty proud of the model’s text-rendering capabilities, which is really useful for things like advertising,” Brichtova said. “If you want a product somewhere, or even just a slogan, it needs to be accurate … We definitely anticipate filmmakers and other kinds of creators are going to be using this model as well.”

The more professional use cases might be better served by the Omni Pro model, which should perform better across all Omni tasks. Google hasn’t said when it will release Pro yet, but Brichtova said that will happen when “we feel like we’re at a point where we have a step change above Flash.”

Catch up on the rest of Google IO 2026’s big news

Google Search as you know it is over

Google updates Gemini app to take on ChatGPT and Claude

Google introduces Gemini Spark, a 24/7 agent assistant with Gmail integration

How to use Google’s new information agents

With AI image generators widely available online and more sophisticated than ever, it’s never been harder to tell if an image is authentic. But on Tuesday, OpenAI announced two new measures to help fight the problem.

The company has committed to an open standard called C2PA, which adds a clear signal in metadata that an image was generated by AI. OpenAI is also partnering with Google to include an invisible watermark called SynthID, which will be harder to detect, but also harder to erase if bad actors try to cover their tracks.

The new protections only apply to images generated by OpenAI products, so they won’t affect the flood of imagery coming from less reputable AI tools; they can help ensure that OpenAI isn’t part of the problem.

OpenAI is also previewing a public verification tool that will check for both signals, allowing users to easily test whether an image was generated using AI. Initially, the tool will only extend to images generated by OpenAI products; the company hopes to expand it to cover other tools over time.

Founded in 2021, the Coalition for Content Provenance and Authenticity (C2PA) is a non-profit dedicated to mitigating the harmful effects of AI imagery on public discourse. The C2PA standard has been adopted by a range of Google products, but adoption remains inconsistent across the industry. Because the C2PA signal is clearly accessible in the metadata of each file, it can be manipulated, and is most useful among trusted users.

SynthID is a newer effort designed to be a more robust measure to meddling. Developed by Google, the SynthID watermark is designed to persist even when bad actors attempt to remove it, either through screenshots, resizing or digital manipulation.

The two systems are meant to complement each other, with each addressing the other’s weaknesses.

“Watermarking can be more durable through transformations like screenshots, while metadata can provide more information than a watermark alone,” OpenAI noted in its announcement. “Together, they make provenance more resilient than either layer would be on its own.”