Google I/O 2025: As Google pushes AI Mode in Search, will it help ease antitrust pressure?
Google's experimental AI Mode feature in Search, that enables deeper and more complex search queries, is now rolling out to all users in the US to start with. The announcement came during the opening keynote of Google's annual I/O developer conference held in Mountain View, California on Tuesday, May 20, and is focused on showing off the company's latest advancements in AI, Android, augmented reality, and more.
Over the coming weeks, users in the US will start to see a new AI Mode tab in Google Search and as a button in the search bar within the Google app.
While Google on Tuesday drew the curtains back on several new AI products, including the company's most advanced image and video generation models, AI Mode in Search was the big news to come out of I/O 2025 for two main reasons.
Google has long dominated the search engine market, shaping the structure of the internet through the rise and fall of content driven by its search engine. But this dominance has come under threat recently with the rise of AI chatbots that have emerged as a viable alternative to looking up information online.
In this context, AI Mode is billed to be the future of Google Search as it moves away from showing users a list of blue links on a web page to providing an end-to-end search experience. However, Google is also facing a prominent antitrust challenge where the US government is pushing for the company's search business to be broken up after a district court ruled last year that Google had an illegal monopoly of the online search market.
AI Mode is powered by a custom version of Google's Gemini 2.5 frontier model. 'AI Mode is where we'll first bring Gemini's frontier capabilities, and it's also a glimpse of what's to come,' Elizabeth Reid, vice president and head of Google Search, said in a blog post.
Searching in AI Mode is different from the typical search on Google as users can dive deeper into the web by asking follow-up questions and finding hyper-relevant content that matches their search query, according to Google. In order to display more helpful responses with web links, AI Mode relies on a query fan-out technique, where AI breaks down the search query into sub-topics and further initiates a multitude of search queries on a user's behalf.
At I/O, Google also announced new capabilities of AI Mode that will become available to select, approved users under Google Labs in the coming weeks and months.
One such feature is Deep Search, where AI Mode uses the same query fan-out technique to issue hundreds of searches and gather information across sources to generate an 'expert-level' research report with citations. This entire process will take minutes, Google claimed.
With Search Live, users can have a back-and-forth vocal conversation with the AI chatbot about what they see on their phone screen or through their phone's camera. AI Mode will also come with certain agentic capabilities. For instance, when a user asks to book tickets or make restaurant reservations in AI Mode, it will kick off a query fan-out process by scanning several sites and presenting the user with various options.
AI Mode will also provide a better shopping experience for users by showing them products and product details in an easy-to-browse panel, along with a virtual try-on feature, AI agent-driven price tracking, and autonomous checkout. This suggests that Google is going toe-to-toe with challengers like OpenAI which recently rolled out its own shopping update to ChatGPT.
In addition, AI Mode will also have the capability to analyse datasets and real-time information to generate graphs, charts, and other data visualisations.
During the Google antitrust search remedies trial, AI prominently featured in the arguments made by both sides. The US Department of Justice (DOJ) essentially argued that Google's new AI search products were just another way to lead users to its search engine and extend its dominance of the online search market. It has also warned that Google could come to dominate the AI space, the same way it has dominated search, if the court did not take any action.
Google, on the other hand, has argued that there is growing competition in AI by pointing out the successes of OpenAI's ChatGPT and others. It called witnesses like Eddy Cue, an Apple executive, who suggested that AI was shrinking the antitrust threat posed by Google in search. Cue's testimony, however, tanked Google parent Alphabet's stock by eight per cent after he revealed that search volume to its Safari browser (running on Google's search engine) had declined for the first time in 22 years.
To reassure shareholders over the threat posed by AI chatbots, Google quickly issued a statement saying that overall queries in Search have continued to grow, except that users are accessing it in different ways.
At Tuesday's I/O event, Google's announcements largely echoed these arguments. It re-emphasised that 'there's been a profound shift in how people are using Google Search' with more than 1.5 billion people using visual search via Google Lens every month. 'In our biggest markets like the US and India, AI Overviews is driving over 10% increase in usage of Google for the types of queries that show AI Overviews,' it said.
However, the accuracy of answers provided by AI-based search products are still under question as large language models (LLM) are prone to hallucinations. There are also concerns that Google may further eat into click-through rates as it leans more heavily on AI-generated search results. This would be bad for publishers and content creators who rely on traffic from Google Search.
In terms of privacy, AI search engines could increase user tracking to gather more detailed feedback in order to improve search results. For instance, AI Mode has a personalised context setting which enables it to pull information from other connected Google apps like Gmail to show tailored results. However, Google has said that this will be an opt-in feature and users can choose to connect or disconnect the apps with AI Mode at any time.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Time of India
17 minutes ago
- Time of India
Enabling the last mile with AI generated insights
In a country like India, enabling effective last-mile customer experiences is crucial. When a customer walks into a retail outlet, they must have a positive interaction, one that ultimately leads to conversion. The question is, how can AI help drive better last-mile conversions? To explore this, in association with has released an episode of 'Enabling Phygital Experiences : Empowering the Last Mile with AI-Generated Insights', featuring industry leaders from the marketing world. This episode delves into how AI can transform last-mile delivery by equipping sales representatives with real-time insights, simulating customer interactions , and improving pitch precision. Featuring voices from Harman India , Signify, and it highlights AI's role as a value multiplier when applied thoughtfully across sales, marketing, and customer support. The episode features Akhil Sethi, head of digital marketing, Harman India, Anmol Reen Jaloota, head of the digital enabling function, Signify Greater India, and Arun Subramaniam, co-founder, Jaloota observed, 'At the ground level, sales representatives need to be able to sell better. They need to understand the unique selling points of a product or project and how to effectively pitch it to the customer. AI helps answer the fundamental question: What do I sell to this customer? By analysing past purchases, financial background, affluence, and even geographic preferences, AI helps identify customer needs, whether someone would want a fan or not, for instance.' He added that AI is already playing a significant role across various functions at Signify, saying, 'We've been using AI in lighting design. Imagine large offices or open spaces where someone is selling lighting solutions but doesn't know the best placement. We've introduced an agentic AI lighting designer that suggests optimal lighting layouts.' AI is not just a tool, it is a powerful mechanism that helps sales representatives better understand their customers, gain deeper insights into products, and ultimately sell more effectively. That is the real value AI brings to the table. Elaborating further, Sethi said, 'What we've observed is that AI is advancing at a very rapid pace across every department. In customer support and marketing, we use AI to track conversations happening both in the social media space and offline, using various tools at our disposal. We conduct sentiment analysis to understand how different products are performing, looking at ratings, reviews, and customer feedback. On the marketing front, AI not only helps us target specific cohorts more effectively from an engagement perspective but also plays a crucial role in driving conversions and sell-through.' One of the biggest challenges in last-mile engagement is that every conversation is unique, no two are ever the same. Unlike marketing, where you often have access to rich data or assumptions about the customer, sales often operates with far less information. Counterintuitive as it may seem, a salesperson typically knows less about a customer walking in than a marketer does, even at an aggregate level. Equipping sales reps with relevant data the moment a customer enters is easier said than done. Subramaniam noted, 'Need analysis is what most customers ask for, yet it is often underused. Why? Because sellers on the field typically lack the confidence to ask personal questions like "How many people are in your family?" or "What's your salary bracket?" Budget-related queries are easier, but deeper profiling becomes uncomfortable. So, how can we support sellers with AI without demanding too much input from them? We see AI helping in two ways. First, by giving sellers a tool, like a ChatGPT-style assistant, they can use it to get quick answers. While possible, it may not be well received during live customer interactions. Second, and more effectively, AI can simulate real-world sales scenarios. We use AI to conduct role plays, sending virtual customers with different personalities, needs, and behaviours to train the seller. Some may be polite, others more difficult. This helps salespeople prepare and adapt, improving performance without direct pressure during a live pitch.' This episode explored how AI can enhance last-mile delivery, highlighting both practical use cases and potential pitfalls. When applied in the right context, AI serves as a powerful value multiplier, capable of simulating diverse customer interactions and helping sales representatives better prepare, personalise, and perform.
Mint
31 minutes ago
- Mint
BP Puts AI at the Heart of Its Efforts to Boost Performance
(Bloomberg) -- When engineers prepare to drill for oil, they know the spot deep underground where the well must end and can choose their starting point, but there are many possible routes in between. Optimizing that subsurface path — evaluating geological opportunities and challenges to ensure a successful job — has been a time-consuming task for engineers. Now, through BP Plc's technology center in Houston, a new AI-powered tool is dramatically streamlining the process and running thousands of scenarios to determine the best trajectory. You may be interested in 'It basically takes the time it would've taken people to do that from months down to days,' said BP Executive Vice President of Technology Emeka Emembolu. The technology 'is a massive game changer and it's getting us better outcomes in the wells we're drilling.' Artificial intelligence is being used by many companies across the oil industry. Exxon Mobil Corp. deployed the technology to help develop its flagship offshore discovery in Guyana. Autonomous drilling has played a role in productivity improvements seen in the US shale industry. The potential for this technology to deliver significant gains in operational efficiency has particular relevance for BP. Under pressure from unhappy shareholders and aggressive activist investor Elliott Investment Management, the company is seeking to reverse a long period of poor performance by boosting growth and profitability. After several years of focusing on clean energy, oil drilling has renewed importance as BP pivots back toward fossil fuels. The financial targets that underpin Chief Executive Officer Murray Auchincloss's strategy reset all require doing more with less — curbing capital expenditure, cutting costs, raising returns and giving more cash to shareholders. To help achieve these goals, BP is pushing AI into every part of its operations, Emembolu said in an interview in London, where he will be speaking at the Tech Week conference. 'Our technology agenda is central to growing oil and gas, central to helping us focus our downstream business and to invest in the transition with discipline,' Emembolu said. The drilling optimization tool is already being used in fields from the Gulf of Mexico — a key driver of US oil output growth this year — to Azerbaijan, where BP earlier this month advanced a $2.9 billion natural gas project. In the Permian Basin of West Texas and New Mexico, where BP plans significant investments in shale oil production, Emembolu said an AI-generated 'morning report' is directing field hands to sites most urgently in need of work and reduce the amount of time spent driving across the sprawling basin's roads. Near Chicago, where BP's Whiting refinery processes large volumes of crude from Canada, the company is working with Palantir Technologies Inc. to embed data engineers to optimize processes on site to reduce costs and improve operational uptime. Disruption at the facility can have a significant impact on BP's earnings, such as in the first quarter of 2024 when a storm led to a lengthy shutdown. The technology is also being used outside of core oil and gas businesses — identifying optimal locations for the fastest electric-vehicle chargers; helping Indian motorists avoid lines at fuel stations with mobile notifications; advising German convenience store managers on how many pastries to bake each morning. 'In terms of costs, we're looking at things from all scales,' Emembolu said. 'Nothing is too big or too small for us to look at.' More stories like this are available on
Time of India
36 minutes ago
- Time of India
Sam Altman's eye-scanning identification tech expands to UK
Synopsis Tools for Humanity, co-founded by Sam Altman, is launching its eye-scanning Orb devices in the UK, starting in London and expanding to other major cities. This global expansion aims to provide a unique digital identity, World ID, to combat AI-driven identity theft and misinformation.
