Google AI: How search, Gmail will change to compete with ChatGPT

In the past, Google has offered programs such as Google News Showcase to compensate publishers without undermining its core argument that the copyright doctrine of fair use permits use of their material. The company has also been more open to striking deals with wire services such as the AP, which are in the business of licensing content. As you’re having the conversation, you’ll find links right on your screen if you want to dig deeper into your search. Users will be able to access the feature by opening the Google app and tapping the new “Live” icon to ask their question aloud. They will then hear an AI-generated audio response, and they can follow up with another question. Google has been working on significant UI updates to the Translate app to better accommodate its upcoming AI-driven features like Insights, Regional variants, Ask a Question, and Practice.

Project Mariner

In July, Cloudflare Inc., a web infrastructure and security company, announced a “pay per crawl” program that lets creators bill AI services for access to their content. Google’s licensing project is tailored to specific products, according to another person familiar with the plan. They didn’t share additional details of the program, such as possible terms. “We’ve said that we’re exploring and experimenting with new types of partnerships and product experiences, but we aren’t sharing details about specific plans or conversations at this time,” a Google spokesperson said in a statement. Google didn’t immediately respond to questions on when new AI features would be available in Search.

Once created, your Gems appear in the sidebar, and you can chat with them just like regular Gemini. Because of how Deep Research works, it’s likely to be more accurate than typical AI searches — though as with anything AI, you’ll want to double-check important information and be on the lookout for hallucinations in the results. An APK teardown helps predict features that may arrive on a service in the future based on work-in-progress code.

• Beyond practical usage, you can also use these generative AI text capabilities for all kinds of cool creative applications.
• There’s also the 1 million token context window, which is equivalent to 1,500 pages of text or 30,000 lines of code.
• As you’re having the conversation, you’ll find links right on your screen if you want to dig deeper into your search.
• Magic Eraser lets you remove unwanted objects and people from photos with a simple tap, while Photo Unblur sharpens those frustratingly blurry shots you can’t retake.

Google Search AI Mode

Gemini Live turns your phone into a conversational AI companion that you can actually talk to, and it’s also free with the Gemini app on iOS and Android. As of this writing, Google says Gemini Live is “Available to mobile users in 45+ languages and over 150 countries.” The AI handles different art styles too — specify if you want something photorealistic, cartoon-style, or even painted in the style of a famous artist. Just note that content limitations mean it won’t create images of real people or inappropriate content. The quality isn’t quite as good as some AI image generators — I’ve found OpenAI’s GPT-4o to be the best so far — but it’s still very capable for most uses.

What features you get with Google AI Pro and AI Ultra July 2025

If you also happen to use Google Docs, Google plans to bring similar capabilities there using what it’s calling Duet AI for Workspace. Silicon Valley and the media industry have been at odds over tech companies’ use of news content to build AI programs, with publishers worried it will further erode their relationship with readers. In late 2023, the New York Times sued OpenAI, alleging that the startup and its largest investor, Microsoft Corp., relied on copyrighted articles to train the popular ChatGPT chatbot and other artificial intelligence features. As mentioned, all of these capabilities and many more in other Google applications like Maps, Photos and Universal Translator are being enabled by a technology called generative AI. This is a relatively new type of artificial intelligence that doesn’t just analyze and respond to data but can generate new information based on an input request using sophisticated data models. As the world’s most popular search engine, Google serves up information and links in response to billions of queries every day.

Earlier this year, Google announced a partnership with the AP to provide news for its Gemini chatbot, its first deal of that nature. The tech giant has also explored an audio AI news product under its Gemini brand that leverages the content it’s licensing from the AP, according to a person familiar with the matter. Google cites articles and online outlets in its AI Overviews, which are short, AI-generated responses that top many search results. While publishers believe those summaries have cut traffic to their websites, they have been hesitant to shield their content from Google’s AI tools for fear of hurting their visibility in the company’s search results.

Google ups the ante on AI to compete with ChatGPT. Here’s how search and Gmail will change.

However, it is possible that such predicted features may not make it to a public release. Additionally, you get the “Highest access” to Gemini 2.5 Flash, 2.5 Pro, Deep Research, and Audio Overviews. There’s also the 1 million token context window, which is equivalent to 1,500 pages of text or 30,000 lines of code. In comparison, free users get a 32K context window (“around 50 pages of text”).

Surprisingly, Google, which has long been known as a leader in AI technology, seemed to initially be caught off guard by these developments. That was particularly the case when Microsoft announced a big partnership with OpenAI to enable a new generative AI-powered version of their Bing search engine (see “New Bing with ChatGPT brings the power of AI to Microsoft’s signature search engine” for more). What’s particularly ironic is that many of the core developments that led to generative AI – particularly a concept known as a transformer model – were initially created by Google researchers. When it comes to searching, the process of “Googling” something is about to undergo the most dramatic changes you’ve ever seen. One of the most impressive capabilities of the technology underlying generative AI is the ability to summarize large amounts of information in a very concise manner. Practically speaking, what that means is it can take data from multiple sources and stitch it all together in a way that gives you exactly what you want in a single response.

The company plans to launch a pilot project initially with about 20 national news outlets, according to one of the people, who works at a media outlet that was contacted and asked not to be identified discussing private talks. Google is rolling out the ability for users to have a back-and-forth voice conversation with AI Mode, its experimental Search feature that lets users ask complex, multi-part questions. With the new Search Live integration, users can have a free-flowing voice conversation with Search and explore links from across the web. Beyond practical usage, you can also use these generative AI text capabilities for all kinds of cool creative applications.

But that’s not the only part of the app getting a UI refresh, as we’ve now spotted more changes slated for Google Translate, this time targeting the Conversation mode. NotebookLM is available for free, though it does have usage limits — you can create up to 100 notebooks, add 50 sources per notebook, and generate 3 Audio Overviews per day. Upgrading will basically expand on those limits, though even the paid plans still have limits. The feature is rolling out integration with Google’s other apps too, so you can discuss plans and have Gemini automatically create calendar events, or ask about restaurants and get real-time info from Google Maps. While there are paid plans for Google AI, aka Gemini, you can access many of its most popular features for free.