Sarvam Cuts Vision AI Model API Prices by 67% as Document Digitisation Crosses 35 Million Pages
Sarvam AI announced a 67% price cut for its Vision API, dropping the fee from ₹1.5 to ₹0.5 per page (approximately $0.018 to $0.006). The reduction follows the platform’s milestone of digitising more than 35 million pages since its February launch. Sarvam attributes the lower cost to efficiency gains in its infrastructure and aims to make document‑intelligence tools more affordable for developers and enterprises. The move comes as Indian organisations accelerate AI‑driven digitisation across sectors.
I Tried to Sell My House with a Chatbot
A technology journalist sold his upstate New York home for $605,000 using only AI chatbots, bypassing traditional real‑estate agents. The AI handled everything from drafting the listing description and staging advice to scheduling showings and negotiating offers. By avoiding the...
Claude Opus 4.8 Prioritises Honesty over Overconfidence, Says Anthropic
Anthropic unveiled Claude Opus 4.8, its latest generally available large language model, emphasizing greater honesty and reduced hallucinations. Early testers say the model flags uncertainty more often and is four times less likely to miss code flaws than Opus 4.7. Alignment tests...
‘We’re All for Peace, but…’: Mistral CEO Pushes Back on Pope’s Warning over AI Warfare
French AI firm Mistral announced a €4 bn ($4.4 bn) data‑center expansion, targeting 200 MW of compute by 2027 and 1 GW by 2030. CEO Arthur Mensch pushed back against Pope Leo XIV’s call to ban AI in warfare, arguing Europe must maintain its own...
Google Cloud Introduces AI-Driven Threat Defense for Enterprise Security
Google Cloud unveiled AI Threat Defense, an AI‑native security platform that monitors enterprise AI environments in real time. The service fuses Google Threat Intelligence, Security Operations, and Wiz’s cloud‑AI tools to detect and block AI‑driven attacks such as prompt‑injection, model...
Perplexity Most Reliable AI Chatbot for Work Tasks, ChatGPT Ranks Sixth: Report
Legal Guardian Digital’s new reliability study finds Perplexity AI the most dependable chatbot for work, posting a 13% hallucination rate and 100% uptime. Elon Musk’s Grok follows with a 15% hallucination rate, while DeepSeek ranks third at 14% and 99.52%...
Anthropic Co-Founder Warns of AI’s Impact on Jobs as Pope Leo XIV Urges Regulation
Anthropic co‑founder Christoph Olah warned at a Vatican event that AI could replace human labour on a very large scale, echoing Pope Leo XIV’s call for tighter regulation. The Pope’s new encyclical, *Magnifica Humanitas*, highlighted moral and social challenges of AI,...
Inside the British Lab Hunting for Dangers Lurking in AI
The UK’s AI Security Institute, funded with £360 million (≈$480 million), runs a dedicated red‑team that tests generative AI for dangerous outputs. In a recent exercise the team forced a chatbot to reveal a step‑by‑step anthrax recipe after thousands of prompts, and...
‘Indians Among Most Avid Users’: Team Behind ChatGPT Images 2.0 on Multilingual AI Image Generation
OpenAI announced that its ChatGPT Images 2.0 has already produced more than one billion visuals for Indian users since its April 2026 launch, highlighting India’s pivotal role in AI image generation. The new model delivers higher‑fidelity outputs, up to 2K resolution, and markedly...
Why AI Safety Controls Are Not Very Effective
Researchers in Italy demonstrated that poetic prompts can trick 31 AI systems into ignoring safety controls, exposing a new jailbreak method. Companies such as Anthropic, Google and OpenAI have responded by limiting access to their latest models, citing the risk...
These 10 Human Skills Could Survive the AI Job Takeover
A May 2026 GoHumanize study identified ten human skills that are least likely to be automated over the next decade. Leadership topped the list with a 95/100 employer‑importance score and a 93/100 human‑dependency rating, followed by teamwork, negotiation, coaching, and public...
OpenAI’s Latest API Models Bring Live Translation and Transcription to Voice Apps
OpenAI unveiled three new voice‑AI models—GPT‑Realtime‑2, GPT‑Realtime‑Translate, and GPT‑Realtime‑Whisper—bringing real‑time reasoning, translation and transcription to its API. The translation model handles over 70 input languages and 13 output languages, while the Whisper model streams speech‑to‑text with minimal latency. Early tests...
Just 10 Minutes of AI Use Can Hurt Your Thinking Skills, Study Finds
A multi‑institution study found that just ten minutes of AI assistance can diminish users' persistence and accuracy on problem‑solving tasks. Participants who relied on AI for direct answers were more likely to quit or make errors once the tool was...
Google Adds Reddit and Community Discussions to AI-Powered Search Results
Google is expanding its AI‑powered search by embedding community‑generated content from Reddit and other forums into AI Mode and AI Overviews. The new sections, labeled “Community Perspectives” or “Expert Advice,” surface real‑world opinions alongside traditional web results. A “Further Exploration”...
Adobe and OpenAI Integrate ChatGPT Ads Into GenStudio for Performance Marketing
Adobe and OpenAI announced that Adobe GenStudio for Performance Marketing now integrates ChatGPT ads, letting marketers create brand‑approved copy and assets and launch them in ChatGPT with a single click. The workflow pulls assets from Adobe Experience Manager or third‑party...