Latest AI News
YouTube Expands AI Deepfake Detection Tool to Politicians, Officials & Journalists
The platform is giving public figures new ways to flag and remove impersonation videos.
View
OpenClaw Bans Perplexity’s GitHub Account Over ‘Sloppy’ Pull Requests
“I expect better from them than sending an untested slop PR and thinking that’s okay and something we want to promote,” said Peter Steinberger.
View
Josh Talks Launches Full-Duplex Conversational AI Model in Hindi
Josh Talks has open-sourced the Hindi duplex conversational model to encourage further research on conversational voice AI for Indian languages.
View
ChatGPT Introduces Interactive Visual Learning Tools for Maths and Science
OpenAI’s ChatGPT will enable users to explore maths and science concepts through real-time dynamic diagrams.
View
Uber Says AI Now Writes 31% of Its Codebase
Uber revealed that around 11% of pull requests are opened directly by its autonomous agents.
View
Sonata Software Files Bankruptcy Petition in the US Against Client Over $10.64 Million in Dues
The company says Q4 FY26 PAT will be higher than Q3 despite the dispute, and that revenue will remain in line with February 6 guidance.
View
Inside the “AI Cold War” Between TCS and Infosys
While TCS is focused on large compute infrastructure, Infosys is building AI systems designed for enterprise operations.
View
Google DeepMind Releases Gemini Embedding 2 in Public Preview
The model allows developers to perform retrieval, classification and semantic search across different types of media using a unified representation of data.
View
Google Expands Gemini in Chrome to India With Support for 8 Indic Languages
Google announced the expansion of Gemini in Chrome to India on Wednesday. The artificial intelligence (AI)-powered assistant for the browser was first released in select markets in September 2025, and now the company is expanding it to more regions. The Gemini extension in Chrome appears as a side panel on the right side of the browser and can answer any questions about the webpage. The Mountain View-based tech giant has also added support for eight Indic languages with the India release.
View
Google brings Gemini in Chrome to India
Google announced Wednesday that it is bringing Gemini integration for Chrome to new regions, including India, Canada, and New Zealand. The rollout will let users use Gemini in Chrome through a sidebar on desktop, enabling them to ask Google’s AI chatbot questions about the content on the screen, get information from their Gmail, Keep, Drive, and YouTube, and compare tab contents. As part of the new rollout, Gemini will also support languages including Hindi, Bengali, Gujarati, Kannada, Malayalam, Marathi, Telugu, and Tamil, in addition to English and Chrome’sother newly supported languages. Google first introduced Gemini in Chrome in the U.S. through a floating window last September. The company introduced sidebar-based Gemini tools earlier this year. Users who get access to this feature will see an “Ask Gemini” icon on the tab bar, which they can activate for any tab and ask questions, summarize content, or create a quiz to understand a topic. Google said that Gemini can also work across tabs. This means you can mention multiple tabs to get an answer, which is helpful when you are comparing items to shop for or tickets to buy for a trip. Gemini can also connect with different tools, get your information, and give more personalized answers. It can connect to Gmail, Maps, Calendar, YouTube, and other Google apps for contextual answers, too. For instance, you can directly compose an email using Gemini in the sidebar on Chrome and send it to someone without leaving the window. You can also ask Gemini to summarize a YouTube video and list the main points alongside timestamp markers. The assistant can also schedule meetings or brief you about your day. Loading the player… Users can also use Google’s Nano Banana 2 generative AI tool directly in Gemini for Chrome to transform images. For instance, you can upload a photo of your room while buying furniture and ask the assistant to transform the image to see how an item would look in the room. The company said that, along with desktop, it is also rolling out Gemini support in Chrome for iOS in India. When available, the option will show up in the address bar through a page tools icon. Google in January launched increasedagentic capabilities, which can take over your browser and complete tasks on your behalf, for U.S.-based AI Pro and AI Ultra users. The company is keeping this function out of the latest expansion for users in India, New Zealand, and Canada.
View
AI-powered apps struggle with long-term retention, new report shows
With the top app stores flooded with AI apps, developers may think the best bet for turning a profit is to integrate artificial intelligence technology into their own products. However, a new study focused on the subscription app ecosystem across iOS, Android, and web is calling that assumption into question. RevenueCat, a company that offers subscription management tools used by over 75,000 app developers, said in its2026 State of Subscription Apps Reportthat AI integration is not a guarantee of long-term retention. Instead, AI-powered apps struggle to retain subscribers, with people canceling their annual subscriptions — a metric known as churn — 30% faster than non-AI apps, at the median, according to the report. The report is based on an analysis of the subscription app providers that useRevenueCat’s toolsto manage their more than 1 billion in-app transactions, generating more than $11 billion in revenue for developers annually. As one of the more popular tools in this space, its data represents a healthy sample in terms of trend analysis. Among the many interesting findings, the report noted that most of the apps using the company’s platform are not yet powered by AI. AI-powered apps account for 27.1% of apps across all categories, compared with 72.9% for non-AI apps. Still, it’s a growing category, as roughly one in four apps is now AI-powered. (To be clear, the AI-powered apps category includes the popular AI chatbots, like ChatGPT and Gemini, as well as any app that markets itself as being AI-powered.) Photo & Video apps have the biggest share (61.4%) of AI-powered apps, while gaming has the smallest share at 6.2%. Travel (12.3%) and Business (19.1%) are also low-AI segments. The more surprising figures are around AI apps’ ability to retain their paying customers. AI apps underperform on retention at both a monthly and annual level, RevenueCat’s data shows. Annual retention, a metric focused on the app’s ability to retain subscribers after 12 months, was 21.1% for AI apps, compared with a higher 30.7% for non-AI apps. Monthly, AI apps saw 6.1% retention rates versus 9.5% for non-AIs — a difference of 3.4 percentage points. The only area where AI led on retention was on the weekly front, where AI apps had 2.5% retention rates compared with 1.7% for non-AI apps. It’s worth noting that weekly subscriptions are not the most popular option for AI apps. These metrics could be influenced by the rapidly changing state of AI technology, which could see users hopping between different AI apps more quickly, as they try to find the one that has the most current technology under the hood. As customers experiment with a growing number of AI apps, they’re also more likely to find that some don’t meet their needs. The report notes that AI apps have 20% higher refund rates (4.2% vs. 3.5% at the median) than non-AI apps do. The upper bound of refund rates for AI apps is also higher (15.6% vs. 12.5%), suggesting there’s “greater volatility in realized revenue and deeper issues in user value, experience, and long-term quality,” the report notes. There are some benefits to being in the AI-powered apps cohort, the data indicates. RevenueCat found that AI apps convert users from trials to paid customers 52% better than non-AI apps (8.5% vs. 5.6% at the median), and AI apps monetize their downloads around 20% better than non-AI apps (2.4% to 2% at the median). AI apps also generate 39% or higher monthly realized lifetime value (RLTV), a metric that measures the actual net value of an average paying user over time. AI apps’ median on this metric is $18.92 per month, compared with $13.59 for non-AI apps. AI apps also sustain a 41% or higher RLTV on an annual basis, at $30.16 vs. $21.37, also at the median. The overall takeaway from the report’s findings is that AI can drive strong, early monetization, but these apps are struggling to sustain their value with customers over time.
View
Amazon launches its healthcare AI assistant on its website and app
Amazonannouncedon Tuesday that it’s expanding access to its healthcare AI assistant to its website and app. The assistant, called Health AI, was previously only available on the app for One Medical, the healthcare companyAmazon acquired for $3.9 billion in 2023. Health AI can answer questions, explain health records, manage prescription renewals, book appointments, and more. Users don’t need to be Prime subscribers or One Medical members to use the assistant, Amazon says. Amazon says that while Health AI can answer general health questions without having access to an individual’s medical information, it’s designed to serve as a personalized health assistant that can offer more tailored guidance and take actions, such as connecting you with healthcare professionals and treatments. Of course, there are risks associated with sharing your health information with AI, andresearchersarewarning againstdoing so, cautioning that companies are pulling user conversations for training. Amazon said in its announcement that it trains “Health AI models on abstracted patterns without directly identifying information.” For example, if multiple patients ask about medication interactions, Amazon says it might use these patterns, while keeping patient names private, to improve how Health AI responds to similar questions. The company also said that all interactions with Health AI happen within a HIPAA-compliant environment, and that conversations are protected by “encryption and strict access controls.” However, Amazon didn’t share exactly how conversations are encrypted and who has access to them. TechCrunch has reached out for specifics. With a user’s permission, Health AI gets access to their health information through the Health Information Exchange, which is the nationwide secure system for sharing patient medical data. Health AI can then interpret your lab results, diagnoses, and medical records to provide accurate, personalized answers about your symptoms and medication, Amazon says. If you need professional care, Health AI can connect you to a One Medical provider. Prime members in the U.S. using Health AI receive up to five free direct-message care consultations with a One Medical provider for over 30 common conditions, including cold and flu, allergies, acid reflux, pink eye, UTIs, erectile dysfunction, anti-aging skin care, hair loss, and more. Non-Prime members can connect with One Medical providers through Amazon’s pay-per-visit option. Users can sign up for Health AI on theAmazon Health page. As Amazon expands access, users will receive an email once they are able to access the assistant. Once users have access, they need to create a or sign in to their personal Amazon Health profile. They can then start a conversation by typing their health question to Health AI on Amazon.com or in the Amazon app. Users can ask questions like “Can you explain my recent cholesterol results and what they mean for me?” or “I’m feeling congested and have a sore throat. What should I do?” Amazon’s expansion of Health AI comes as popular AI services have quickly moved into the healthcare space. In January, OpenAI releasedChatGPT Health, a version of its chatbot tailored to answer health questions. A week later, Anthropic announced its own healthcare-focused product,Claude for Healthcare.
View
