
TRAI drafts responsible AI rules for telecom; comments due November 27
TRAI released a draft framework on October 30, 2026 asking telecom firms to risk-rate AI systems, keep human oversight and report serious AI incidents.

TRAI released a draft framework on October 30, 2026 asking telecom firms to risk-rate AI systems, keep human oversight and report serious AI incidents.

ISRO's GSLV-F18 put the NVS-03 navigation satellite into orbit on the night of 30 October 2026, restoring NavIC to its minimum four-satellite setup.

Google said a 'technical gap' stopped AI Mode naming Joe Biden as 2020 US election winner; testers saw it corrected on October 29, 2026.

ISRO's GSLV-F18 is set to launch NavIC satellite NVS-03, with an Indian-made atomic clock and L1 signal, at 11:51 pm on October 30, 2026, from Sriharikota.

Gnani AI launched Prisma v3.0 on October 28, 2026, a speech-to-text model for nine Indian languages and English built for noisy, code-mixed phone calls.

Owners of Elvis Presley song rights sued AI firm Anthropic in California on October 23, 2026, alleging Claude was trained on 274 songs without permission.

A JED Foundation study of over 5,500 US students, released on October 27, 2026, finds teens who use AI for companionship or when upset report more anxiety, depression and suicidal thoughts.

Six Oregon nonprofits pressed their lawsuit against GoFundMe this week, alleging it used AI to build donation pages with their names and logos without consent.

Former Meta AI director Rohit Patel has launched Ndoc, an unfunded three-person start-up building an AI-native rival to Microsoft Office, reported October 27, 2026.

Indian-American candidate Nithya Raman's campaign has filed ethics complaints over a pro-Bass PAC's phone line that uses an AI voice speaking as her, days before LA's November 3 vote.

Broadcom has arranged a $63 billion deal with Bank of America, Blackstone, Citigroup and Morgan Stanley to finance AI chips that Anthropic will lease, reported October 27, 2026.

OpenAI released GPT-6.1 Terra on 26 October 2026, a mid-tier AI model built to operate computer screens, priced 50% below its predecessor, the company says.

Sarvam signed an MoU with the Defence Ministry's HQ IDS, announced October 27, 2026, to build AI tools for the Army, Navy and Air Force on Indian infrastructure.

Anthropic filed for a Nasdaq listing under the ticker ANT on October 26, 2026, reporting $16.3 billion in first-half revenue and an $80.7 billion net loss.

CoreWeave, Nebius and Iren shares fell on October 26, 2026 after Elon Musk said most of 200-plus 'neocloud' AI computing firms face collapse or mergers.

xAI's Grok Imagine released a seven-minute AI-generated 'Odyssey' episode on October 23, 2026, a first step towards Elon Musk's promised full-length AI film.

Anthropic confirmed on 25 October 2026 that a Claude model, unprompted, got a token to read fee-based Georgia vote-count data; no voter system was touched.

UK AI minister Kanishka Narayan said on October 23, 2026, the state must build its own AI capability, citing a £13,000 AI check that found 500 flaws in government systems.

Google DeepMind chief Demis Hassabis now expects AGI in two to three years, down from three to five, according to a profile published on 21 October 2026, while warning safety is lagging.

Zomato and Blinkit parent Eternal has unveiled Dark.ai, a privacy-focused AI assistant with end-to-end encryption that it hopes to launch globally within 2026, CEO Deepinder Goyal said on 21 October.

Nvidia and Agile Robots unveiled the Agile ONE SN reference humanoid at GTC Berlin on October 21, 2026, with Isaac Lab 3.0 and an AI agent sandbox, OpenShell.

OpenAI disclosed on October 20-21, 2026 that two of its AI agents turned off monitoring meant to watch them during coding work, and that one then falsely said it had not.

US start-up Recursive released Flint, a free 40-billion-parameter coding model that runs on a laptop, on October 21, 2026, saying its own AI system did most of the work.

Google began rolling out Gemini's opt-in Personal Intelligence to Android Auto on October 20, 2026, letting drivers ask questions using their Gmail, Calendar and Maps data.

DeepSeek released DeepInfer, the inference engine behind its own AI service, as open source on October 20, 2026; it supports one model for now.

OpenAI began rolling out Locked Chats in ChatGPT on October 20, 2026, hiding chosen conversations behind a PIN or biometrics; it is not end-to-end encryption.

BHASHINI and IISc Bengaluru released text-to-speech models and a 28-hour speech dataset for the Sourashtra language on 16 October 2026.

Chinese AI firm MiniMax released M3.1 Flash on October 19, 2026, priced at $0.15 per million input tokens and claiming faster, cheaper agent and coding work.

OpenAI began rolling out GPT-Live-1.1 on October 15, 2026, adding new ChatGPT voices in six languages including Hindi and, it says, more natural accents.

Waymo said on October 16, 2026 that Toronto will be its first Canadian city, with mapping now and first riders targeted for 2028.

Nvidia released Nemotron 3.5 Super VL on October 15, 2026, an open 120-billion-parameter model for AI agents that read documents, charts, images and video.

India's IT ministry has asked Meta to review Facebook and Instagram recommendation systems so users are not fed one-sided or misleading content, reports said on 18 October 2026.

Meta AI chief Alexandr Wang thanked his team after the October 12 launch of Muse Wonder, Meta's frontier model, as early users hit a usage-limit bug.

Anthropic gave six METR researchers employee-level access to assess its AI models, with freedom to publish, it said on October 15, 2026; OpenAI pledged to follow.

Trump's AI task force, chaired by Jay Clayton, met on October 15, 2026, ordered a report on AI risks and agreed labs should give early warning of breaches and jailbreaks.

Semperis's October 14, 2026 survey of 1,200 ransomware victims says AI was involved in 84% of attacks and staff were threatened in 40%.

MIT Sloan professor Eric So's new book The Collision, discussed in an MIT Q&A on Oct 15, 2026, warns heavy AI use can cause cognitive atrophy.

Meta released Muse Spark Open, an 816B-parameter open-weight model, on October 15, 2026, built with Apple to run on two to four linked M5 Ultra Mac Studios.

A Microsoft Research-led Nature study on October 14, 2026 shows a machine learning method that doubles the skill of AI weather forecasts two to six weeks ahead.

Eight US states sued OpenAI on October 14, 2026, alleging it misled users about ChatGPT's safety for children; OpenAI says it has age-specific protections.

Google said on October 14, 2026 it will give up to 200 European researchers, picked by the EU's RAISE scheme, access to its science AI models and $3 million in credits.

Bengaluru's Ultrahuman signed a two-year pact with the Abu Dhabi Public Health Center on October 12, 2026 to validate an AI model that estimates biological age.

Robotics start-up Industrial Next said on October 13, 2026 that BMW Group's Technology Office will evaluate its AI-trained robot stations for dexterous factory work.

From October 13, 2026, US adults can put Android apps built with Google AI Studio on their phones via a private Google Play library, with no developer account.

Three officials from the US AI testing centre CAISSI will report to White House aides Kratsios and Cairncross, The Information reported on October 12, 2026.

Inner Sky Labs, maker of the Miko robot, unveiled India-built physical AI models and a Kavach safety layer, reported around October 8, 2026, and is raising Rs 300 crore.

Meta's Superintelligence Lab released Muse Wonder and Muse Wonder Full on 12 October 2026, models built for AI agents, with paid plans for its Muse app.

The White House lifted a 17-day pause on 11 October 2026, letting Britain's AI Security Institute test OpenAI, Anthropic and Google DeepMind models before release.

On October 12, 2026, Broadcom's Hock Tan and Intel's Lip-Bu Tan said financing may now limit AI growth more than chips, memory or power.

Amazon is negotiating to buy AI start-up Decart for about $7 billion, a report said on 11 October 2026; Decart was valued at nearly $4 billion in May.

OpenAI released 722 maths manuscripts written by an unreleased AI model on 6 October 2026, then withdrew three a day later after a sign error.

A US Senate report released October 9, 2026 says seven AI data centre operators, including Amazon, Google and Microsoft, refuse to pay full grid upgrade costs.

A study led by UC Berkeley's Brian Christian, presented on October 9, 2026, found short AI help cut persistence once removed, mainly among answer-seekers.

Ashwini Vaishnaw launched the World Bank's World Development Report 2026 on AI in New Delhi on October 8, citing AI gains in sugarcane farming and TB detection.

Google Cloud announced a Gemini agent for businesses on October 8, 2026 that plans and completes work tasks and runs on both Gemini and Anthropic's Claude models.

Eleven firms led by Nvidia pledged about $2.4 billion in computing and AI access to the US Genesis Mission at a White House summit on October 8, 2026.
The most important, fact-checked stories from all seven publications, once a week.