At its I/O developer conference in Mountain View, Google unveiled what it called the biggest Search upgrade in over 25 years. The redesigned interface accepts text, images, documents, video and open browser tabs and responds with synthesized answers rather than a ranked list of links. Alongside it, Google launched persistent AI agents in Search that monitor topics and push notifications without being prompted.

From Keywords to Continuous Intent

One year after its debut, AI Mode has surpassed 1 billion monthly users, queries more than doubling every quarter, Google noted. VP of Search Liz Reid told reporters query volume hit an all-time high last quarter. Google Search and advertising revenue reached $60.4 billion in the first quarter, up 19% year over year. That ran counter to predictions AI-generated answers would hollow out Search’s ad business, PYMNTS found.

The redesigned interface anticipates intent, expanding as users describe what they need. Users can attach documents, images, videos and Chrome tabs to the search box, The Verge reported.

Agents That Work Without Being Asked

The more consequential announcement for commerce and payments is what Google calls information agents. Unlike tools that respond only when prompted, these agents operate continuously, monitoring topics and pushing notifications when something relevant surfaces.

Use cases range from apartment hunting to tracking product drops and financial developments, Google said. TechCrunch noted that information agents are the next evolution of Google Alerts, now built on a reasoning layer that synthesizes conflicting perspectives. They launch first for Google AI Pro and Ultra subscribers this summer. That’s the same rollout model governing Gemini Spark, its new personal AI agent running 24/7 on Google Cloud, integrating with Gmail, Docs and third-party services over MCP.

Commerce Infrastructure Underneath

The company also introduced Universal Cart, built on Google Wallet, working across Search, the Gemini app, YouTube and Gmail. Once a product is added, it monitors for price drops, surfaces price history and alerts users when an item is back in stock.

The interface is new. Google’s Agent Payments Protocol creates a verifiable link between user, merchant and payment processor, with tamper-proof mandates ensuring the agent acts within user-specified boundaries, the company explained. The Universal Commerce Protocol, an open standard Google debuted earlier this year with Amazon, Shopify and Walmart, gives AI assistants a common language for any merchant without bespoke integrations.

Universal Cart arrives in Search and the Gemini app in the U.S. this summer, YouTube and Gmail to follow, with UCP-powered checkout expanding to Canada, Australia and the U.K.

What Else Was Announced at I/O

• Google launched Gemini Omni, a multimodal model that accepts image, audio, video and text input and outputs editable video grounded in real-world knowledge. Available today for paid Gemini subscribers, it’s rolling out to YouTube Shorts Remix and the Create app.
• Ask YouTube lets users search the full YouTube catalogue using natural language and follow-up questions. It’s available now for Premium subscribers in the U.S. and rolls out broadly this summer.
• Samsung and Google debuted the first Android XR audio glasses, designed with Warby Parker and Gentle Monster, with Gemini integration and Maps support. They arrive this fall.
• Docs Live, a voice-driven document creation tool built on Gemini Live, lets users speak documents into existence. It rolls out this summer for Google AI Pro and Ultra subscribers on Android and iOS.

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.

The new venture is designed to provide “efficient data center capacity, operations, networking, and Google Cloud’s tensor processing units (TPUs) as a compute-as-a-service offering,” Blackstone said in its announcement Tuesday (May 18).

Customers will have the option to access cloud TPUs along with using them through Google Cloud, the announcement added.

“This joint venture with Blackstone helps meet growing demand for TPUs, which are optimized specifically for efficiency and performance in the AI era,” said Thomas Kurian, Google Cloud’s CEO. “Together, we’re accelerating AI transformation and providing more options for organizations to access accelerated compute capability.”

Google’s TPUs are custom chips optimized for training and inference in advanced AI models, and are used by outside customers such as AI labs and other firms. The custom chips are also used to power Google’s in-house AI offerings like Gemini. 

Under the terms of this new partnership, Blackstone will commit $5 billion to start, with the new company expecting to bring the first 500 megawatts of capacity online next year.

“Google will supply hardware, including TPUs, as well as software and services to the new company so it can rapidly accelerate to meet the growing demand for accelerated computing, leveraging the benefit of Google’s technical and domain expertise,” the announcement added.

A report on the new company by the Wall Street Journal notes that industry observers have for years wondered whether Google would commercialize its TPUs for widespread usage. Google has agreements with two outside companies, that report added: one with Anthropic giving the AI startup access to about one million of its chips, and another with Meta.

In other AI news, PYMNTS wrote last week that while discussions of the technology tend to center around questions of supply, they always ignore the “harder half: demand.”

This came after Anthropic announced an small- to medium-sized business (SMB)-focused AI plugin for tools like PayPal, Intuit, Canva, and Docusign, a move that suggests AI’s next stage could depend less on how much capability technology companies can offer and more on whether ordinary companies can sustain demand inside everyday work.

“After all, the question facing the market is no longer simply whether frontier models can perform astonishing tasks,” PYMNTS wrote.

“It is whether accountants, nurses, insurance adjusters, teachers, procurement managers, financial analysts and their peers across Main Street can integrate AI into the highly specialized workflows they understand better than any Silicon Valley engineer.”

“This reimagined experience offers a suite of powerful capabilities designed to help you better understand the financial world,” the tech giant said in a blog post Monday (May 11).

According to the post, the offering includes artificial intelligence (AI)-powered research, charting tools, real-time intelligence, and the ability to follow corporate earnings calls lives while also getting AI-generated insights.

Google launched Google Finance in the U.S. last August and India in November, and said last month that it planned to expand the tool to 100 other countries, including Australia, Brazil, Canada, Indonesia, Japan and Mexico.

This is happening amid increased, though sometimes cautious, use of artificial intelligence as a financial tool, a trend chronicled by PYMNTS Intelligence late last year.

“AI adoption inside finance today is highest where rules are clear and outcomes are measurable,” said in that report, titled “CFOs Are Letting AI Into the Finance Function, Carefully.”

“Nearly half of CFOs already rely on the technology to monitor cash positions, audit readiness, anomaly detection and compliance oversight. These functions benefit from automation because it reduces manual effort without needing judgment calls that could materially affect financial outcomes,” the report continued.

Chief financial officers are becoming comfortable with AI assisting with judgment-driven analysis, as long as final authority stays with humans. More than half said they would accept AI-generated recommendations on liquidity and payment timing. Upwards of 40% of CFOs said they trust the technology to handle audit logs, integrity alerts and variance analysis.

“This signals a shift from automation to advisory support,” PYMNTS added. “However, willingness declines sharply when recommendations require coordination across multiple systems. CFOs want AI to frame decisions, not execute them when operational risk is high.”

Google Finance’s latest expansion follows last week’s announcement of Google’s new AI- powered bidding and budgeting tools for Search and Shopping.

“By pairing smarter bidding strategies with flexible budgets, you can capture high-value opportunities without the manual heavy lifting,” Josh Braverman, group product manager, Google Ads, wrote in a blog post announcing the tools.

One tool currently in beta is journey-aware bidding, which lets advertisers track the path from leads to sales, thus helping Google AI understand complex lead generation customer journeys.

“By seeing the whole picture, Google AI can better predict what works and optimize your performance,” Braverman said in the post.

“By pairing smarter bidding strategies with flexible budgets, you can capture high-value opportunities without the manual heavy lifting,” Josh Braverman, group product manager, Google Ads, wrote in a Thursday (May 7) blog post announcing the tools.

One tool currently in beta is journey-aware bidding, which enables advertisers to track their full lead-to-sales journeys, thereby helping Google AI better understand complex lead generation customer journeys.

“By seeing the whole picture, Google AI can better predict what works and optimize your performance,” Braverman said in the post.

In another upcoming offering, Google will extend its Smart Bidding Exploration capability to Performance Max and Shopping campaigns. Smart Bidding Exploration helps advertisers capture less obvious queries that they didn’t anticipate and weren’t able to win before. This capability was introduced to Search campaigns last year and has delivered an average of 27% more unique converting users.

“Soon you’ll be able to use that same capability to reach new customers and capture additional conversions across both Performance Max and Shopping campaigns,” Braverman said in the post.

Google will also upgrade its budget pacing in Search and Shopping Campaigns, enabling it to better predict consumer demand and follow demand automatically. These upgrades will be rolled out in the coming months.

“With demand-led pacing, Google AI will better optimize spend to follow consumer demand — capturing more demand on peak days and reducing spend on slower days — all while never going beyond your monthly budget and daily spending limits,” Braverman said in the post.

PYMNTS reported April 29 that executives at Google parent company Alphabet said during an earnings call that AI-driven Search queries are at an all-time high and that more than 30% of search ad spend now uses AI-enabled campaign tools. This reflects how advertisers are adjusting to conversational and context-rich queries, according to the report.

“AI is boosting our ability to deeply understand user intent for a given search query and to find the most relevant ad,” Alphabet Chief Business Officer Philipp Schindler said during the call. “Even when we don’t have a direct user query, we’re making significant strides in improving relevance.”

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.

Outlining the new features in a Wednesday (April 29) blog post, the company said Google TV users can turn ideas into images by visiting the Gemini tab and selecting “Create” to access Nano Banana; generate custom videos from scratch or from an existing image by visiting the Gemini tab and selecting “Create” to access Veo; and search Google Photos with their voice with the help of Gemini.

In addition, Google TV users can change their photos to look like a watercolor or oil painting by using Google Photos Remix; turn their Google Photos album into a dynamic slideshow by visiting the screensaver settings under “Quick Settings”; and watch a personalized feed of short videos by accessing a new “Short videos for you” row on the Home page.

The availability of these features will vary by device, country and language, according to the post.

“Whether you’re bringing hilarious ideas to life, reliving your favorite moments from your Google Photos library or kicking back to watch snackable clips on the big screen, these updates are designed to help you connect, create and share laughs together,” Michael DelGaudio, UX manager and design lead at Google TV, and Gaurav Chaula, product manager at Google TV, said in the post.

In another recent move, Google said April 21 that it embedded two new AI features into Chrome, putting Gemini inside the purchase journey before shoppers reach a buying decision. Skills for Gemini lets users save AI prompts and replay them on any webpage with a single click, while an update to AI Mode in Chrome means clicking a search result now opens a merchant’s page in a side panel alongside an active Gemini session.

On April 8, Google said the new AI-powered Google Finance was set to expand beyond the United States and India, adding more than 100 countries over the following weeks. The new Google Finance offers AI-powered research that answers questions and provides links for additional information, new charting tools, a revamped news feed, and expanded data for commodities and cryptocurrencies.

The demands will be part of the Commission’s effort to get the tech giant to allow other AI assistants to use the same voice activation, search tools and other Android features that Google’s own Gemini assistant uses, according to the report.

If Google does not comply with the demands when they are issued, the company could face a formal European Union probe that could result in financial penalties, the report said.

When the EU announced in January that it was launching an effort to pressure Google to remove barriers to other AI assistants, the company said it was concerned that such a move could “compromise user privacy, security and innovation,” per the report.

In the January announcement, the European Commission said it opened two formal specification proceedings aimed at clarifying how Google must meet key obligations under the Digital Markets Act (DMA). The Commission said the move is intended to guide the company in adjusting its business practices while ensuring that competitors are able to operate on equal terms within the digital marketplace.

One of those cases focuses on Google’s duties under the DMA to offer free and effective interoperability to third-party developers for hardware and software features managed through its Android operating system, the Commission said, adding that the current discussions center on technical functions used by Google’s own AI tools.

The Commission said in the January announcement that it expected to send Google its preliminary findings and outline proposed measures intended to ensure compliance with the DMA within three months and then complete the proceedings within six months.

In another, separate case, the European Commission announced April 16 that it released proposed measures it wants Google to implement to make its data available to rival search engines. The Commission said that its goal is to enable third-party search engines to optimize their search services and contest Google’s position in the search market.

The company announced these developments Wednesday (April 22) in conjunction with its Cloud Next event in Las Vegas.

Google’s new Gemini Enterprise Agent Platform provides a system for building, scaling, governing and optimizing agents. It builds upon the company’s existing AI development platform, Vertex AI, by combining model selection, model building and agent building capabilities with new features for agent integration, DevOps, orchestration and security, according to a Wednesday press release.

The company also introduced three new agents in Google Security Operations to help organizations defend against the malicious use of AI. These include a Threat Hunting agent that searches for novel attack patterns and stealthy adversary behaviors, a Detection Engineering agent that identifies coverage gaps and creates new detections for threat scenarios, and a Third-Party Context agent that enriches workflows with contextual data from third-party content, per a Wednesday press release.

Google Cloud’s new $750 million fund will provide resources and incentives to global consulting firms, systems integrators, software providers and channel partners to help their joint customers adopt agentic AI. Resources available to partners will include AI value assessments, Gemini proofs-of-concept, Gemini Enterprises practice building, agentic AI prototyping and development, Wiz security assessments and usage incentives, according to a Wednesday press release.

The new agreement with Thinking Machines Lab will see Google Cloud provide the AI startup with additional AI infrastructure capabilities and capacity, including A4X Max VMs with Nvidia GB300 GPUs as well as services such as Kubernetes Engine, Spanner, Cluster Director, Cloud Storage and Anywhere Cache. Myle Ott, founding researcher at Thinking Machines Lab, said in a Wednesday press release that this infrastructure got the company running “at record speed.”

Sundar Pichai, CEO of Google and Alphabet, said in a Wednesday blog post that the pace of technological change has never been faster than it has been over the past year.

“Our first-party models now process more than 16 billion tokens per minute via direct API use by our customers, up from 10 billion last quarter,” Pichai said.

Skills for Gemini lets users save AI prompts and replay them on any webpage with a single click. An update to AI Mode in Chrome means clicking a search result now opens a merchant’s page in a side panel alongside an active Gemini session, inserting the AI into the browsing session the moment a shopper shows intent.

The deployment sits on top of considerable infrastructure. Chrome runs on roughly 3.5 billion devices and holds about 65% of the global browser market, according to StatCounter. Gemini counts 750 million monthly active users across its surfaces, Google said, up more than 100-fold in two years.

Skills live inside the Gemini side panel built into Chrome desktop. A user writes a prompt, saves it, then activates it on any page by typing a forward slash or a plus. It runs on the current tab or across however many tabs the user picks.

Google launched Skills with a curated library spanning Shopping, Recipes, Budgeting and Productivity. Most launch examples target retail directly: spec comparisons across open tabs, review summaries and gift recommendations scored against a budget. Once saved, a workflow travels with the user into every product page they open.

AI Mode works from the other end. When a user finds a result worth exploring, clicking it opens the merchant’s page in a side panel inside the same session. Gemini reads the page and the broader web at once and fields follow-up questions against both, Google said.

The merchant’s page is no longer where discovery ends. It’s where Gemini starts working.

Google isn’t stopping at the browser. The Gemini app is now available on macOS as a native desktop experience, reachable from anywhere on the desktop with a keyboard shortcut.

Anthropic is moving in a parallel direction inside productivity software: Claude for Excel and Claude for PowerPoint now let users save workflows as one-click Skills that repeat across open files, targeting finance and enterprise teams working across spreadsheets and decks.

Proposed measures announced Thursday (April 16) by the European Commission, the EU’s competition regulator, would let other search engines see the tech giant’s search data, including ranking, query, click and view data.

The goal is to let third-party online search engines, or “data beneficiaries,” optimize their search services and contest Google’s position in the search market, the commission added.

“Data is a key input for online search and for developing new services, including AI,” European Commission Executive Vice-President for Clean, Just and Competitive Transition Teresa Ribera said in a news release.

“Access to this data should not be restricted in ways that could harm competition. In fast-moving markets, small changes can quickly have a big impact. We will not allow practices that risk closing markets or limiting choice.”

According to the release, the proposed measures, under the EU’s Digital Markets Act (DMA), cover issues such as the eligibility of data beneficiaries to receive search data, including that of artificial intelligence (AI) chatbots that have search functions.

The measures also cover the scope of the data Google must share and the “means and frequency” by which Google must share it, as well as measures to make sure personal data is anonymized, the commission added.

“Hundreds of millions of Europeans trust Google with their most sensitive searches — including private questions about their health, family, and finances — and the Commission’s proposal would force us to hand this data over to third parties, with dangerously ineffective privacy protections,” Clare Kelly, Google’s senior competition counsel, said in a statement to PYMNTS.

“We will continue to vigorously defend against this overreach, which far exceeds the DMA’s original mandate and jeopardizes people’s privacy and security.”

Google also argued that it had already launched a DMA-compliant anonymized dataset for rivals to use its data and expertise to build competing search products, and that this solution was already working.

The commission says it will collect input on the proposal before imposing a final decision by July 27.

Google is also reportedly dealing with scrutiny from the commission into whether the company had distorted ad auction mechanisms in a manner that may have driven up costs for advertisers. The company is facing fines of at least 9.5 billion euros ($11.2 billion) due to alleged antitrust issues in the region.

Launched in the U.S. in August and India in November, the AI-powered Google Finance will soon be rolled out in Australia, Brazil, Canada, Indonesia, Japan, Mexico and other countries, with full local language support, Google said in a Wednesday (April 8) blog post emailed to PYMNTS.

The new Google Finance offers AI-powered research that answers questions and provides links for additional information, new charting tools that let users toggle technical indicators such as moving average envelopes and candlestick charts, a revamped news feed, and expanded data for commodities and cryptocurrencies, according to the post.

In addition, for corporate earnings calls, the new Google Finance offers live audio, synchronized transcripts and AI-generated insights, per the post.

The latest availability of the new Google Finance can be checked at Google’s Help Center.

When Google announced in August that it was testing a new version of Google Finance in the U.S., the company said the new AI-powered features were joining the real-time market quotes, international exchanges, financial news and analytics already offered by the site.

The company announced the expansion of the new Google Finance to India in November, saying it would launch there with English and Hindi support.

Google also said in November that it was updating Google Finance to include Deep Research that allows users to type open-ended financial questions and receive AI-generated responses that include citations and links; enhanced charting that offers technical indicators and historical overlays; and data from prediction markets Kalshi and Polymarket, which provide market-based probabilities for outcomes such as inflation rates, GDP growth and interest rate decisions.

Management at Google parent company Alphabet reported during a February earnings call that the company is seeing sustained demand for AI compute across its consumer products, enterprise platforms and cloud infrastructure.

The company said Gemini was processing more than 10 billion tokens per minute via direct application programming interfaces (APIs) in the fourth quarter, up from 7 billion in the previous quarter.