Latest AI News
AI Job Apocalypse or Productivity Boom? Silicon Valley Can’t Decide
As tech layoffs mount, AI leaders are increasingly divided on whether the technology will eliminate jobs, create new ones, or simply change the nature of work.
View
Lovable signs multiyear deal with Google Cloud to up usage 5x, source says
Lovable and Googleannouncedan expanded multiyear collaboration on Wednesday. Lovable, the fast-growing Stockholm vibe-coding startup, has long been a Google Cloud user. Under the new agreement, it will be a much bigger one. While the companies did not disclose the dollar figure, a person with knowledge of the deal tells TechCrunch it involves a fivefold increase in Lovable’s footprint on Google Cloud, including AI usage. As part of the deal, this individual tells us, Lovable will gain expanded access to both Anthropic’s Claude — the AI model widely used for coding tasks — and Google’s own Gemini models. The Anthropic piece in particular is interesting. Google invested $10 billion in Anthropic in cash and compute credits in April, promising another $30 billion if Anthropic hits certain performance targets. It made that investment at a $350 billion valuation — just one month before Anthropic raised a staggering$65 billion roundthat valued the company at nearly $1 trillion. This deal stands to help Anthropic hit those targets, because Lovable is one of Europe’s fastest-growing startups on record. According to Lovable, it crossed $400 million inannualized revenue in February, having added $100 million in a single month with just 146 employees. The company claims that more than half of Fortune 500 companies use its product in some fashion. The deal also plugs Lovable into several other parts of Google’s ecosystem. Lovable’s new agent will be available through Google Cloud’s enterprise agent marketplace, the Gemini Enterprise Agent Gallery — an arrangement the two companiesfirst telegraphedat Google’s major U.S. cloud conference in April. And to help secure the code that both humans and agents write, Lovable will integrate with Wiz, Google’s biggest ever acquisition at $32 billion, whichofficially closedin March, a year after it wasannounced. The integration will allow Wiz to identify and remediate security problems in real time. By selling Lovable’s agents through Google’s marketplace, the cloud giant says enterprise procurement and billing will be simplified, making it easier for Lovable to land more enterprise customers. The calculus for Google is simple enough. If it can keep both Lovable and Anthropic growing by attracting deep-pocketed enterprises, the revenue helps fund the $180 billion to $190 billion in capital expenditures Google plans to spend this year. The company is already in the process ofselling a record-breaking $85 billion in equityto cover some of that, so only another $100 billion or so to go.
View
Google’s Dreambeans, its weirdest-named AI tool to date, will turn your life into a cartoon
Google Labs, the tech giant’s team devoted to experimental product design, has launched a new AI-fueled app for iOS and Android that will quite literally animate your life. Behold,Dreambeans. Why is it called that? We’ll get to that later. First, what is it? Gozde Oznur, the product lead behind the new app, told TechCrunch that the idea is to use data culled from across your various Google services to generate a curated list of AI-illustrated “stories.” These stories come in a variety of different shapes and forms, although — in general — they seem to be lifestyle suggestions. Oznur describes them as “places to visit, topics to explore, things to try, upcoming trips, events that you should be aware of.” Dreambeans generates these ideas based on a user’s Google data. “With your permission, Dreambeans uses Personal Intelligence to connect information from Google apps like Gmail, Calendar, Photos, YouTube and Search History, to curate a finite collection of daily stories designed to spark new ideas,” the company says. So for instance, some stories may be geographical recommendations — like suggesting a new coffee shop near where the user lives that they might be interested in. Or, as is the case in thismarketing video, if you’re getting a new dog and that event has been marked in your Google Calendar, Dreambeans might deliver some insights about what it’s like to live with a new puppy. Still other stories may simply be news articles curated from the web, based on a user’s past interests. Oznur said the app has also been built as a doomscrolling antidote, in that it only provides users with a limited number of stories per day — typically 10 to 14. The idea is to get a few inspirational ideas and then go out and live your life, she said. A lot of companies are currently trying to court the user that is sick of phone addiction. I recentlyreviewed a startup, Bond,which also uses AI to auto-generate lifestyle suggestions for the user. What about privacy protections? According to Oznur, they are pretty solid. The only person with access to the app’s stories is the user, she said. Users can also delete their data whenever they want, and can choose which Google services they want to connect to the tool. Finally, where did the name “Dreambeans” come from? The idea for the name was generated, in part, by the way the system works while you are asleep, she said. “The dream part is literal, because while you sleep, the app is working through everything across your connected apps, because, as you can imagine, it’s a lot of data that it is distilling,” Oznur said. “The beans part is about how you kind of start your day with a freshly brewed cup of coffee. It has processed everything overnight and hands you a concentrated drop of inspiration in the morning.” Dreambeans is currently only available for eligible U.S.-based Google AI Ultra subscribers on Android and iOS. However, there is also awaitlistthat is available to users with a personal Google account.
View
Alphabet’s record-breaking $85B raise for Google’s AI business is a helluva good signal
If Alphabet’s record-breaking $85 billion stock sale signals investor appetite for AI-related offerings — and it does — we can safely say that investors are voracious. Google’s parent company had initially intended to sell a first tranche of $40 billion worth of various equity instruments — two different classes of shares, plus smaller “depositary shares” priced to be accessible to a broader range of investors. But the offering was so oversubscribed that it raised $45 billion instead, CEO Sundar Pichai said in apost on Xon Monday. Among the buyers: Berkshire Hathaway, still known for its love of value investing, picked up $10 billion worth. Alphabet plans to sell another $40 billion worth next quarter, for $85 billion total. Even $80 billion would have topped the record for equity offerings previously set by Brazilian oil producer Petroleo Brasileiro SA, which raised $70 billion in 2010,Bloomberg reports. Now, it’s true that these investors are buying shares of Alphabet, not shares in a younger, possibly debt-riddled AI startup. Alphabet is a very healthy business: $110 billion in revenue (with high profit margins) in Q1 alone, up 22% year-over-year. Still, the money from this stock sale is earmarked for AI. “Part of our multi-year investment strategy to meet the AI opportunity ahead and support the demand we’re seeing from enterprises and consumers,” as Pichai described it. At Google I/O last month, hesaidthe company expects to spend between $180 billion and $190 billion on capital expenditures — largely on AI infrastructure and data centers — before the year is out. The timing matters beyond Alphabet itself. As Anthropicgets ready to go public, this enormously successful stock sale is a very good sign for the broader AI IPO pipeline. It indicates that public investors, particularly the deep-pocketed institutional ones, are ready to pony up. The upcoming SpaceX IPO is expected to smash records for cash raised and valuation, and Anthropic’s deal is expected to do the same, possibly surpassing SpaceX. OpenAI is also waiting in the wings. But all of this rests on public investors’ appetite — not just private VCs — remaining strong, and then staying that way. An unprecedentednearly $8 trillion in AI spendinghas been committed over the next five years. That money has to come from somewhere — and that somewhere includes individual company revenues, loans, and capital raised through stock sales. Whether public markets have the stomach to absorb that much, for that long, is the question that every AI company eyeing an IPO should be thinking about right now.
View
Publishers will be able to opt out of AI Search, thanks to new regulation
The U.K. has just imposed legal guardrails on Google’s AI search onslaught. On Wednesday, Googleannouncedcompliance with the U.K.’s regulatory requirements, which state that the tech giant must offer publishers a way to opt out of being aggregated into AI search. To opt out, publishers will be able to use a new toggle in Google’sSearch Console, a free service that allows website owners to manage their web presence in Google’s search results. Once opted out, the publisher’s site will not be shown in Google’s generative AI Search features, like AI Overviews, AI Mode, or AI Overviews in Discover. (Google, of course, makes a point to note in the same announcement that its AI Overviews now have over 2.5 billion monthly active users, and its AI Mode has surpassed one billion monthly users.) The tech giant says it will initially test the opt-out option with a subset of U.K. publishers before rolling it out globally. The U.K.’s Competition and Markets Authority (CMA)calls the moveto put publishers back in control of how their content is used a “world first,” and points out that it will put publishers, including news organizations, into a stronger position to negotiate content deals with Google for use of their content in AI features. The CMA had first designated Google as having “strategic market status” last October, laying the groundwork for future regulations. InJanuary, it pushed Google to give website publishers a choice as to whether their content is aggregated into AI search features or used to train stand-alone AI models. Alongside the opt-out toggle, Google will also now be required to make sure publisher content in AI features is properly attributed, using clear links. Google suggested that it’s complying with this as well, noting that it had recentlyincreasedthe number of inline links directly within its AI responses, and added website previews to encourage users to click through. Google notes that a website’s decision to opt out of generative AI search features will not be used as a ranking signal for traditional Google search. The company, however, will present new metrics in its Search Console to hopefully sway publishers who could be considering opting out, including impression metrics and other information about which of their pages appear in AI responses, and in which countries. More metrics will be added over time, Google said.
View
These two founders left Goldman and Meta to build voice AI for markets everyone else overlooked
Customer support and service are among the hottest sectors in voice AI right now. But building a product that sounds human and responds without noticeable delay turns out to be much harder in some markets than others — and most of the major players weren’t built with Africa and the Middle East in mind. AethexAI, a startup founded last year to close that gap, has raised $3 million in pre-seed funding led by 4DX Ventures, with participation from Enza Capital, Dorm Room Fund, Mojo Ventures, and Stanford GSB 26 Fund. Individual investors include Stanford faculty, telecom executives, and AI researchers from Anthropic. Rather than using existing orchestration tools likeVapiandLiveKit, the company built its own small model and orchestration layer from scratch to handle the localized dialects of English, French, and Arabic spoken across its target markets — a decision driven, as we’ll get to, by the particular demands of operating in the region. The company is also launching its platform for enterprises to try out its tech and sign up for its services, along with APIs and SDKs for developers to experiment with its models. The startup was founded by Mariama Diallo and Ayooluwa Odemuyiwa. CEO Diallo worked at Goldman Sachs and later joined YC-backed ModelML as a product and growth hire. CTO Odemuyiwa graduated from Caltech, worked at Meta, and enrolled at Stanford Business School before co-founding the company. The pair wanted to build something for emerging markets and started looking for opportunities. Businesses around the world are racing to adopt AI tools to automate parts of their operations. But that doesn’t always work out. In Egypt, a call center automated a significant share of its calls, but rolled the system back because of poor results, the founders found. Several support centers in Africa told them that finding and hiring engineers to automate calls at the right cost was a persistent headache. “The latency and jitter that we saw on automated calls in this region were outrageous. If we had become orchestrators, we might have had to use large models that were hosted outside the region, resulting in higher latency. We realized that in order for this to work, we have to use very small models and cut latency at every step,” Odemuyiwa told TechCrunch about the decision to build the company’s own models and orchestration layer. AI labs that deploy their latest models usually spend millions training them and acquiring data. AethexAI found a solution for both. Rather than chasing the largest possible models, it decided that small models are enough to tackle the latency problem while maintaining accuracy and developed its own Kora series, with parameters ranging from 300 million to 1.7 billion. That’s a fraction of the size of the LLMs, which is precisely the point. To train these models, the startup used anonymized recordings from a call center partner. It also shipped hard drives to radio stations across Africa to collect more audio data. To keep costs down, it built a contributor network of university students to annotate data and pronounce local names. As a result, the startup says, it’s now handling more than 17,000 calls per day. On the business side, the company is taking care to walk clients who are new to voice AI through the process, offering onsite demos and workshops to help them identify the best use cases for automation. “We always tell customers that we cannot be everything for everybody right now. We’re small. When we start talking to a company, we ask them to pick one use case that is the most important to them to start [with],” Diallo said. The startup is open to working across all industries, but at the moment, a big part of its use cases involves calls for debt collection, customer activation, or KYC — Know Your Customer verification, the standard identity-checking process used by banks and telecoms. The company is hiring forward-deployed engineers on a contract basis to serve local markets and building channel partnerships with telecoms providers to handle telephony for voice AI calls. Plug-and-play solutions, it says, simply won’t work here. Walter Baddoo, co-founder and managing partner of 4DX Ventures, argues that the Africa and Middle East market is fundamentally different from the markets most voice AI companies were built to serve. “Enterprises in Africa and the Middle East process roughly three times the call volume of their Western counterparts, as voice is still the dominant channel for customer interaction,” he said. “Incumbent systems were built for Western markets characterized by high-end GPU infrastructure, standard English and European speech environments, and enterprise workflows common in the U.S. and Europe. That creates real gaps when enterprises need systems that handle dialects, code-switching, and informal speech patterns, and that work within their existing telephony infrastructure and their actual price points.” Put another way, while companies like ElevenLabs, Deepgram, Sierra, and Cognigy are expanding globally at a fast pace, the markets they were built for and the markets they are entering aren’t always the same thing. Startups like AethexAI are betting that the gaps — models specialized in local dialects, on-the-ground partnerships, infrastructure built for the region — represent a market opening that the giants have neither the incentive nor the architecture to close.
View
Amazon will show AI product images when you search for some reason
In what may be one of the more questionable uses of AI to date, Amazon announced on Wednesday that it will display AI-generated images of products within its shopping app based on users’ search queries. That’s right — a retailer where people shop for real-world products thinks that displaying fake photos will “help” consumers better find what they’re looking for. Enough already. Here’s how Amazonsays in a blog postthat the feature will work. Customers may have something in mind but don’t know the right term to describe it in a way that returns useful results. (The examples Amazon gives are things like “cowl neck” for a style of shirts or “rattan” for furniture.) When someone enters a search query, they’ll be shown a variety of AI-generated product images below their autocomplete suggestions. (See above photo.) For instance, if you search for a blue gingham dress, you might see a few dress styles — short or long sleeves, varying lengths, and other differences — appear as visual options. The idea is that clicking one would direct you to search results that better match that style, powered by Amazon’s visual search capabilities. In reality, it’s somewhat bananas for a retailer to make up fake products as a way of guiding users to search results. For starters, it’s potentially misleading — customers who don’t read carefully may think they’re being directed to a page where they could find that exact dress, then be disappointed when it isn’t available. And there’s the fairly obvious question of why you’d make up product images when you have a website full of real photographs of real products — which is presumably what an online shopper actually wants to see. The feature follows a number of other attempts by Amazon to integrate AI into its retail site and shopping app, with mixed results. On the more useful end, Amazon alreadysummarizes customer reviewsvia AI, so you don’t have to read them all to get a sense of the key pros and cons of a product. More bizarrely, it last year rolled out ashort audio product summaryfeature in which AI experts describe a product’s highlights, podcast-style. Other recent AI features include AI-generated “shoppable collages” to direct people to curated pages devoted to a particular fashion style;Amazon Lens Live, which scans products in a camera’s view to find visual matches; the ability to add text to visual searches; and a Lock Screen visual search widget for iOS. Earlier this month, Amazon alsoreplaced its Rufus AI chatbot with Alexa for Shoppingto enable natural language shopping queries via voice and text.
View
Everything Microsoft Announced at Build 2026
From Scout and Project Solara to new MAI models and GitHub Copilot upgrades, here are the biggest announcements from Build 2026.
View
Google Rolls Out Gemini Thinking Levels Across Platforms With 'Extended' Thinking Mode for All Users
Google has started rolling out thinking levels for multiple Gemini AI models across platforms, including the Android and iOS apps, a company executive has announced. The new modes are available to all paid and free users in select global markets, including India. The Gemini app now features up to three thinking levels, depending on the Gemini 3.5 and 3.1 series models. The list of levels also includes a new “Extended Thinking” mode, which lets Gemini take more time before providing an AI-generated answer to users. According to reports, this is expected to improve the quality of responses the Gemini app provides.
View
Coralogix raises $200M on bet that someone needs to watch the AI agents
Coralogix, a Boston-headquartered software monitoring startup founded in Israel, has raised $200 million in a new funding round, betting that the rise of AI agents will drive demand for a new generation of tools to monitor, troubleshoot, and manage increasingly autonomous software systems. The Series F financing comes just 11 months after Coralogixraised $115 million in a Series E round, a pace that reflects just how quickly investor appetite for AI infrastructure companies has accelerated. The new round values the startup at $1.6 billion post-money and was led by Advent and the Canada Pension Plan Investment Board (CPPIB), with participation from Greenfield Partners and Brighton Park Capital. The company has now raised a total of $550 million to date. The investment comes as software companiesrace to adapt to the rise of AI agents, software systems that can autonomously write code, investigate problems, and complete tasks that would previously have required a human engineer. Coralogix is among a growing number of infrastructure firms betting that as AI systems move into production, demand will rise for tools that can monitor their behavior, troubleshoot failures, and provide the operational data needed to keep them running reliably. (The more autonomous software you deploy, the more you need to know when something goes wrong and why.) Founded in 2014, Coralogix helps companies monitor the health and performance of software systems by collecting and analyzing operational data such as logs, metrics, and traces — essentially a continuous record of what a software system is doing and how it’s behaving. The platform is used by more than 5,000 customers worldwide, including IBM, Tradeweb, and JFrog, to detect outages, investigate incidents, and optimize applications. The observability industry, where Coralogix competes with the likes of Datadog, New Relic, and Splunk, is being reshaped by the rise of AI. Vendors are increasingly embedding AI into monitoring and incident-response workflows as enterprises deploy more AI-powered applications and agents. The shift is already changing how customers interact with Coralogix’s platform, co-founder and CEO Ariel Assaraf (pictured above, right) said in an interview. More than half of the startup’s enterprise customers now use either its AI agent, Olly, or their own AI models through command-line and agentic interfaces to investigate incidents and query operational data, he said. “The interface layer is slowly getting eroded,” Assaraf told TechCrunch, observing that engineers are increasingly interacting with software through AI assistants and command-line tools rather than traditional dashboards. “Most of the usage is going to be around, ‘How do I connect my LLM to this? How do I operate this through my CLI?’” In plain terms, his customers are less interested in logging into a dashboard and more interested in asking an AI assistant what’s wrong. The shift has coincided with strong growth for Coralogix. The startup grew revenue by more than 60% over the past year and now counts about 30 customers spending more than $1 million annually, Assaraf said, as it expands further into the enterprise market. The company surpassed $100 million in annualized revenue more than a year ago, Assaraf added, though he declined to disclose current figures The startup employs more than 600 people globally, with about 100 based in India, home to its third-largest office after the U.S. and Israel. The India operation, Assaraf said, has evolved into a regional hub supporting customers across Asia while helping Coralogix expand into large domestic enterprises, including financial institutions. Coralogix did not raise because it needed additional runway, Assaraf said, adding that the funding would be used to accelerate investment in AI-focused products, security offerings and global expansion. “In the AI era, execution and speed matter more than any point-in-time valuation,” he said. “We wanted to accelerate, expand, and take a further step into this AI game that we believe we’re leading in our space.” Coralogix does not currently expect to raise additional capital and is working toward profitability over the next few years, Assaraf said. The company is also preparing to operate with the financial discipline of a public company, he said, though he stopped short of committing to a timeline for an initial public offering.
View
Meta’s AI agent for WhatsApp Business is now available globally
For years, WhatsApp has been a communication layer for businesses of all sizes around the world. Meta is now infusing AI into that layer in a bid to turn WhatsApp into a viable piece of workflow software for small and medium businesses. The company on Wednesday said it is making its customer support AI bot, now known as Meta Business Agent, available globally within WhatsApp. The launch comes as Meta has spent nearly two years testing AI agents in WhatsApp Business for customer support in countries like India and Mexico. Meta said the AI agent can answer customer questions, recommend products, book appointments, qualify sales leads, and reroute queries to a person if needed. The company is also making the bot available within Instagram DMs. Meta said it is testing a way for the Business Agent to provide daily briefings of chats that occurred overnight, and provide insights. The company is testing this feature with select accounts on WhatsApp Business, Instagram Pro, Messenger and Meta Business Suite. The company said it’s working to add capabilities like doing market research, highlighting product features, managing users’ calendars, and connecting with tools to extract competitive insights. Meta said it’s also working to enable the Agent to surface businesses when you search for one or share contact details in chat. And, it is building a platform to let larger enterprises create custom agents that can connect to systems like Shopify, Zendesk and Shopee. Meta is planing to charge businesses for using this AI agent by including it in some tiers of its WhatsApp Business Premium subscription. The company noted that large businesses will pay for the agent based on how many tokens they use. This is critical for WhatsApp, which has relied on businesses paying for messaging and click-to-WhatsApp ads.
View
Neo4j to Launch ‘Alternative to Palantir Gotham’ By Acquiring Graph Aware
The graph database company plans to combine GraphAware’s intelligence analysis software with its own graph platform to target government, defence and law-enforcement customers seeking sovereign AI deployments.
View
