Latest AI News

Get the fastest, most accurate AI news every morning

Total articles: 11,301|Added todayToday: 108

Category

Search

Nvidia Opens Run:ai Platform as Open Source

Nvidia released Run:ai’s full platform as open source, two years after acquiring the Israeli-founded company for an estimated $800 million.
Run:ai software manages computing resources for AI workloads and can dynamically share GPU capacity among teams and applications.
The release includes four standalone components; community use needs no paid Run subscription, while enterprise support remains excluded.

calcalistech.comRead more · 1 min →

Today'sEducation

Study Finds AI Homework Gains, Exam Declines

Indian college students report widespread academic AI use, led by research at 84% and writing at 76%.
A preliminary China study found homework scores rose 18% while monthly in-class exam marks fell 20%.
World Bank report cautions findings are not conclusive; Indian colleges cite concerns about overdependence and academic integrity.

India TodayRead more · 2 min →

Today'sSafety & Ethics

AI Labs and Evaluators Clash Over Publication Rights

Anthropic and OpenAI negotiations over independent evaluator publication rights are reported in detail for the first time.
Talks moved from strict editorial control on October 5 to apparent changes in Anthropic discussions by October 8.
METR and Redwood’s six-day review report on the Hugging Face incident included OpenAI redactions.

CNBC TV18Read more · 3 min →

Nilekani Says India Can Lead AI Adoption

At the Bloomberg New Economy Forum on Oct 13, Nandan Nilekani said AI would create opportunities, not widespread unemployment.
Nilekani said India could lead AI adoption by delivering applications to millions, rather than only building powerful models.
He said technology makes up approximately 30% of typical projects and small businesses employing three to five people need not be displaced.

Free Press JournalRead more · 1 min →

Surrey Council Drafts Rules on AI Tools

Surrey County Council committee recommends draft rules restricting Gemini and ChatGPT for council business
Draft policy directs staff to Copilot Chat, Copilot 365 or other council-authorised tools over consumer AI services
Proposed rules allow labelled AI meeting minutes and routine administration, while requiring human oversight in consequential decisions

getsurrey.co.ukRead more · 1 min →

Today'sSafety & Ethics

Scientists Urge International AI Oversight

Around 400 scientists from 34 countries call for an international body to oversee advanced cyber-capable AI.
Open letter proposes industry-wide containment standards and a registry of significant security incidents.
Signatories urge initial network isolation and independent oversight as governments consider formal rules.

Economic TimesRead more · 1 min →

Today'sSafety & Ethics

OpenAI to Watermark ChatGPT Text in EU

OpenAI plans invisible watermarks on eligible ChatGPT and Codex text for EU customers over the coming weeks
The textGrain detector will be limited to approved researchers and expert organizations at launch
TechRadar questions watermark reliability because editing can weaken it and detection can produce false positives

techradar.comRead more · 1 min →

Trending Keywords

Zuckerberg Backs Meta AI Investment

Zuckerberg says Muse Wonder puts Meta back at AI’s frontier and rejects calls to slow development
Meta claims Muse Wonder outperforms Claude Opus 5.5, Fable 5.1, and GPT-6 Astra on select technical benchmarks
Closed Muse Wonder pricing starts at 20 cents per million tokens; premium offering costs $200 a month

NDTV ProfitRead more · 2 min →

OpenAI Explains ChatGPT Dot Access and Costs

OpenAI’s ChatGPT Dots are paid agents that run tasks while users are away from their computers
Pro plans start at $100 per month; Business Premium and approved Enterprise users also get access
Dots connect to more than 4,000 apps, and work they perform counts against monthly plan limits

EngadgetRead more · 2 min →

Today'sSecurity

Anthropic Says Houthis Used Claude for Weapons

Anthropic report says Houthis used Claude to assist weapons development in Yemen
Three programs included a guided rocket and missiles with a range goal above 1,500 miles
Anthropic said safeguards were evaded, but it had no evidence of an operational weapon

dailystar.co.ukRead more · 1 min →

Itoki and Matsuo Institute Analyze 854 Employees’ Work Patterns

Itoki and Matsuo Institute unveiled a model analyzing work patterns of 854 employees.
Mental well-being tended to decline the day after more meetings; it improved for 88% after 60-minute focus sessions.
Sent and received email and chat showed different patterns; the partners plan to expand the model’s reach.

Ledge AIRead more · 2 min →

Singapore Adapts Data Protection for AI

Singapore’s PDPC adapts AI governance through voluntary guidance, industry feedback and collaboration across agencies.
The AI governance framework for agentic AI launched in January and was updated within six months with feedback and case studies.
Singapore’s data protection approach includes ASEAN transfer tools and SEA-LION, built for Southeast Asian languages.

GovInsider AsiaRead more · 2 min →

Champaign County Weighs Data Center Restrictions

Champaign County considers barring NDAs on county business and opposing tax incentives for hyperscale data centers.
The committee was set to discuss both proposals on October 13, with no final action planned.
A moratorium covering facilities with at least 10,000 square feet of processing area is set to end in April.

GovTech AIRead more · 1 min →

Accenture Song Outlines Brand Strategies for Agentic Commerce

Accenture Song explained how companies can prepare for agentic commerce at a media briefing on September 2, 2026.
The two strategies are to standardize product information and build brand-owned AI agents.
JINS AI supports in-store service using records of conversations between store staff and customers.

Impress DIGITAL XRead more · 2 min →

Google Opens Gemini 4 Argon to the Public

Google began general availability of Gemini 4 Argon on October 13 local time.
Subscribers to three paid plans can use Argon; introductory API pricing is $2 per million input tokens and $10 per million output tokens.
Google will offer six months of free access to up to 10,000 researchers and plans to expand Fairwind participation to hundreds of companies.

ITmedia AI+Read more · 1 min →

Today'sSecurity

OpenAI Employee Arrested Over Recording Devices

OpenAI identified an employee who allegedly planted recording devices at its office and reported the matter to the FBI.
The Justice Department charged the employee with attempting to intercept conversations; conviction could bring up to five years in prison and a $250,000 fine.
Two devices were found at the office; OpenAI said it found no evidence that model weights or customer data were taken.

ITmedia AI+Read more · 3 min →

Pew Finds AI Survey Replicas Off by 12.4 Points

Pew Research Center released results on September 30 from AI replicas of thousands of panelists answering nearly 300 survey questions.
Across three surveys, replicas missed responses by an average of 12.4 percentage points; 28% of questions had errors above 15 points.
Pew said AI responses are unsuitable for replacing public opinion polls, citing group stereotypes and limits in models’ knowledge cutoffs.

MobiinsideRead more · 3 min →

Character.ai Files for Chapter 11

Character Technologies filed voluntarily for Chapter 11 while keeping Character.ai running and pursuing an asset sale.
Disney obtained a license to some intellectual property for $20 million plus transition support under a stalking horse bid.
If no buyer commits to continuing the platform, it will close in stages; all services and paid subscriptions remain active for now.

ITmedia AI+Read more · 2 min →

Google Play Adds Private App Runs and Gaming Features

Google Play lets users run Android apps they build without developer accounts.
Ask Play adds app comparisons, while Game Finder launches in English in the US.
Google adds 48-hour Sprints for 2–5 players and Community Posts for all games.

ITmedia AI+Read more · 2 min →

Today'sSafety & Ethics

Half of CISOs Consider Leaving Cybersecurity

ISACA found that 68% of security leaders reported higher job stress over the past five years.
Complex threats were a major stressor for 71% of CISOs, up from 63% in 2025.
Amid the emergence of Mythos-class AI models, 50% of CISOs considered leaving the security industry; leaders outlined seven priorities.

CIO KoreaRead more · 3 min →

Trending Keywords

Last 7 Days