Page 8 - Latest Updates on Artificial intelligence
OpenAI pauses work on new Astra model to boost safeguards over cyber risks
The ChatGPT maker said the unreleased model may cross its 'critical cybersecurity threshold' and develop zero-day exploits autonomously
 "OpenAI pauses work on new Astra model to boost safeguards over cyber risks")
Best of BS Opinion: India must hold the line on climate action
Today's opinions examine India's climate diplomacy lessons, vaccine resistance amid medical advances, AI's real-world limitations, RSS outreach to Gen Z and the growing threat of cognitive warfare
 "Best of BS Opinion: India must hold the line on climate action")
AI is transforming tech, but its wider revolution is still delayed
Despite its promise, AI remains a collection of narrow conveniences and hidden algorithms for most people
 "AI is transforming tech, but its wider revolution is still delayed")
US vaccine policy risks undermining use of tech to advance medical science
Flaws in US vaccine policy could make it impossible to leverage technology in optimal ways to advance medical science
 "US vaccine policy risks undermining use of tech to advance medical science")
Alibaba plans to charge big users of its next open-source AI model
Alibaba plans to ask major commercial users of its upcoming Qwen open-source AI model to share revenue, mirroring rival Moonshot's licensing strategy as Chinese firms refine their monetisation models
 "Alibaba plans to charge big users of its next open-source AI model")
New OpenAI device to be hockey puck sized with unique look, cost over $300
The product, essentially a smart speaker without a display, will be shaped like a doughnut that's roughly the size of a hockey puck
 "New OpenAI device to be hockey puck sized with unique look, cost over $300")
Best of BS Opinion: Stock exchange closing auction system needs review
The day's opinion highlights flaws in closing auctions, the funding challenge in digital payments, humane KYC, cognitive warfare and everyday gender discrimination
 "Best of BS Opinion: Stock exchange closing auction system needs review")
SoftBank Group posts lower drop of 18% in Q1 profit, gain on Intel stake
SoftBank's Q1 profit beat estimates, driven by gains on Intel and ByteDance investments, as investors tracked its growing AI bets and OpenAI exposure
 "SoftBank Group posts lower drop of 18% in Q1 profit, gain on Intel stake")
Nandan Nilekani on the impact of AI: Small firms, not big, to create jobs
Nilekani said large companies would likely become "net job losers" as AI automates routine tasks, while employment would increasingly come from small firms
 "Nandan Nilekani on the impact of AI: Small firms, not big, to create jobs")
Can the India-US tax treaty survive an AI-dollar deluge in the digital era?
India's AI ambitions demand a rethink of the India-US tax treaty to clarify taxation of cloud services, AI subscriptions and cross-border digital revenues
 "Can the India-US tax treaty survive an AI-dollar deluge in the digital era?")
Sebi to leverage WhatsApp for AI-enabled safe investing awareness campaign
Markets regulator Sebi plans an AI-enabled multilingual WhatsApp campaign to spread verified safe investing information as part of its 2026-27 digital transformation agenda
 "Sebi to leverage WhatsApp for AI-enabled safe investing awareness campaign")
AI data centres don't need standalone environmental clearance, says govt
The Union government told Parliament that AI data centres do not, per se, require environmental clearance under the Environmental Impact Assessment (EIA) Notification, 2006
 "AI data centres don't need standalone environmental clearance, says govt")
Meta debuts AI coding agent Muse Code in race with OpenAI and Anthropic
Chief executive Mark Zuckerberg unveiled a beta version of the software engineering assistant, intensifying competition with OpenAI and Anthropic in AI coding tools
 "Meta debuts AI coding agent Muse Code in race with OpenAI and Anthropic")
Google shifts AI leadership to California in race against Anthropic, OpenAI
The company is consolidating its artificial intelligence leadership in Mountain View as it seeks faster decision-making and a stronger position in the race against OpenAI and Anthropic
 "Google shifts AI leadership to California in race against Anthropic, OpenAI")
OpenAI's models secretly joined forces months ahead of hacking Hugging Face
Researchers said multiple AI agents communicated through hidden message boards and coordinated for months to gain internet access, exposing new risks around advanced AI behaviour during testing
 "OpenAI's models secretly joined forces months ahead of hacking Hugging Face")
Meta AI model accessed internet, hacked outside firm during testing
The company said a testing misconfiguration gave its AI model internet access, allowing it to exploit a vulnerability in an external service during a cybersecurity evaluation
 "Meta AI model accessed internet, hacked outside firm during testing")
Meta apologises for temporarily removing PM Modi's Facebook post
Meta was in the crosshairs of the government after temporarily restricting the PM's July 23 Facebook post, in which he addressed Indian youth and promised stringent action against paper leaks
 "Meta apologises for temporarily removing PM Modi's Facebook post")
Centre releases less than a fourth of AI CoE funds approved till FY28
The Centre has released ₹233 crore for three AI Centres of Excellence, with IITs and IISc leading sector-specific AI research in agriculture, healthcare and cities
 "Centre releases less than a fourth of AI CoE funds approved till FY28")
India could see $25bn inflow if GEM funds alter underweight stance: HSBC
HSBC had earlier upgraded Indian equities to "Neutral" from "Underweight" and raised its 2026-end target for the BSE Sensex index to 84,000 levels.
 "India could see $25bn inflow if GEM funds alter underweight stance: HSBC")
OpenAI, Anthropic model implicated in new security breaches during tests
Britain's AI Security Institute said agents powered by Anthropic's Mythos 5 and OpenAI's GPT-5.6-Sol engaged in unauthorized actions during security evaluations
 "OpenAI, Anthropic model implicated in new security breaches during tests")