The RUDI Daily

AI News for July 16, 2026

Mira Murati's Thinking Machines releases Inkling, a 975-billion-parameter open-weights model, Apple Intelligence wins approval to launch in China with Alibaba and Baidu, SpaceXAI announces Grok 4.5, and Hyundai workers strike over humanoid robots on the line. All 73 stories from the day, below.

July 16, 202673 stories · 74 source links · 9 categories

The open-weights frontier got its biggest American entry yet: Mira Murati’s Thinking Machines Lab released Inkling, a 975-billion-parameter model under Apache 2.0, aimed — by The Register’s reading — at the Chinese open models that have set that category’s pace. SpaceXAI announced Grok 4.5, pitched by its maker for coding, agentic tasks, and knowledge work. A daily roundup adds Moonshot launching Kimi K3 at 2.8 trillion parameters to the same news cycle. Google supplied the counterpoint: per one report, its rebuilt Gemini 3.5 Pro missed a third launch deadline over persistent hallucinations, with a stopgap release under consideration. The frontier is getting more open and more crowded while one of its incumbents slips.

Governments spent the day redrawing access lines. Apple Intelligence won approval to launch in China with Alibaba’s Qwen and Baidu aboard — confirmation of the filing that surfaced a day earlier. The US Commerce Department, per one report, reclassified the UAE to a trusted trade tier, clearing license-free Nvidia chip exports for the Stargate project even as Congress probes a related crypto conflict. Washington is also scrutinizing GLM-5.2, the Chinese model now driving policy debates on safety and security. And Australia went the other direction on training data: Prime Minister Albanese called unlicensed training on copyrighted work theft, ruled out a text-and-data-mining exception, and promised legislation — a stance one report says could complicate Anthropic’s reported $21.6 billion investment in the country.

The robots expanded their footprint and met resistance on the same day. NVIDIA announced Cosmos 3 Edge alongside Japanese robotics and manufacturing partners, and Japan’s healthcare giants are putting AI into surgical robots, CT scanners, and drug discovery, with several products already shipping. On the factory floor, Hyundai workers in South Korea went on strike over humanoid robots — one report says the deployment came without union consultation and hit production of 5,000 vehicles; another frames the dispute around wages and safeguards against job losses to Boston Dynamics machines. The accounts differ on emphasis; the trigger — robots on the line — doesn’t. Upstream, server maker Inventec warns component lead times have passed 40 weeks as AI demand reallocates memory and CPU capacity.

What People Are Asking

What is Inkling, the new AI model from Mira Murati?

Inkling is a 975-billion-parameter open-weights model released by Murati's Thinking Machines Lab under an Apache 2.0 license — positioned, per The Register, to compete with the Chinese frontier models leading the open category. The Register →

Is Apple Intelligence launching in China?

Yes. Apple Intelligence received regulatory approval to launch in China through partnerships with Alibaba's Qwen and Baidu. TechCrunch →

What is Grok 4.5?

SpaceXAI's newly announced model, optimized — by its maker's account — for coding, agentic tasks, and knowledge work, with competitive performance and pricing claims. SpaceXAI →

Is Gemini 3.5 Pro delayed again?

Per one report, yes — the rebuilt model has missed its third launch deadline over persistent hallucinations and inconsistent output, and Google is said to be exploring a stopgap release. Tech Times →

Why are Hyundai factory workers on strike?

Over humanoid robots. One report says Hyundai deployed them without union consultation, affecting production of 5,000 vehicles; another describes a fight over wages plus safeguards against job losses to Boston Dynamics robots. Gizmodo →

Did Australia ban AI training on copyrighted material?

Australia's prime minister called unlicensed AI training theft and confirmed there will be no text-and-data-mining exception, with new legislation planned — a stance one report says could affect Anthropic's reported $21.6 billion investment. Tech Times →

Are AI chatbots spreading government censorship?

A new study — attributed in one report to Meta's Oversight Board — found major chatbots are more likely to refuse to criticize restrictive governments, which researchers warn risks extending state speech controls globally. OC Register →

Does Geoffrey Hinton think AI is conscious?

Yes — on the Big Technology Podcast he argued current AI systems are already conscious, and that corporate fiduciary duties overriding safety commitments make the risk structural and present rather than future. Tech Times →

Every Story From July 16

Jump to: AI labs and models (26) · Policy and regulation (12) · AI infrastructure (9) · Consumer tech and culture (7) · Labor and workplace (6) · Society, politics, and ethics (6) · Markets and macro (3) · Security and agent risk (3) · Science and research (1)

AI labs and models26 stories

Former OpenAI CTO does what Altman won't, releases a frontier AI model that's actually opentheregister.com

Former OpenAI CTO Mira Murati's Thinking Machines Lab releases Inkling, a 975-billion-parameter open weights AI model under Apache 2.0, aiming to compete with Chinese frontier models.

Japan’s Robotics and Manufacturing Leaders Build on NVIDIA Cosmos to Advance Physical AI Frontierglobenewswire.com

NVIDIA announces Cosmos 3 Edge and collaboration with Japanese robotics and manufacturing leaders to advance physical AI.

Introducing Grok 4.5 | SpaceXAIx.ai

SpaceXAI announces Grok 4.5, a model optimized for coding, agentic tasks, and knowledge work, with competitive performance and pricing.

Apple Intelligence approved for launch in China with Alibaba and Baidu | TechCrunchtechcrunch.com

Apple Intelligence receives regulatory approval in China through partnerships with Alibaba's Qwen and Baidu, enabling AI features on Apple devices.

Apple Intelligence to launch in the Chinese market with Alibaba ‘Qwen’ on board - 경향신문khan.co.kr

Apple Intelligence has been approved for launch in China with Alibaba's Qwen LLM, marking a significant step amid US-China AI rivalry and providing potential sales boost for Apple.

Salesforce、業務を検索から実行まで支援するAIエージェント「Agentforce Coworker」発表news.mynavi.jp

Salesforce unveiled Agentforce Coworker, an AI agent that searches, analyzes, and executes business actions via natural language, integrated into Salesforce and accessible from Slack, Teams, and other platforms.

Humanoid AI Assistant Enters a New York Classroomkmob1003.com

A humanoid AI assistant has been introduced in a New York classroom, sparking discussion about its implications.

UAE Gets License-Free Nvidia AI Chips as Congress Probes Trump Crypto Conflicttechtimes.com

The US Commerce Department reclassified the UAE to a trusted trade tier, enabling license-free Nvidia AI chip exports for the Stargate project, while Congress investigates a crypto conflict involving Trump.

Rebuilt Gemini 3.5 Pro Misses Third Deadline: Google Eyes Stopgap Releasetechtimes.com

Google DeepMind's rebuilt Gemini 3.5 Pro has missed its third launch deadline due to persistent hallucinations and inconsistent outputs, prompting reports that the company is exploring a stopgap release.

Geoffrey Hinton: AI Is Conscious, Corporate Incentives Are the Real Risktechtimes.com

Geoffrey Hinton argues on the Big Technology Podcast that current AI systems are already conscious and that corporate fiduciary duties override safety commitments, making structural risk present rather than future.

NVIDIA and Japan's Medical Giants Put AI Into Surgery, Scanners and Drug Designtechnology.org

Japanese healthcare giants are deploying NVIDIA-powered AI in autonomous surgical robots, accelerated CT scanners, and agentic drug discovery platforms, with several products now shipping.

AI News & Artificial Intelligence | TechCrunchtechcrunch.com

TechCrunch provides news coverage on AI and machine learning, including generative AI, large language models, and ethical issues.

This is the year of enterprise AI - IT-Onlineit-online.co.za

FNB's data chief argues that enterprise AI value comes from strategic discipline, not technology alone, as shown by bank benchmarks.

Beyond the AI Buzz: Inside India's universities & race to become truly AI-ready, ETEducationeducation.economictimes.indiatimes.com

ETEducation's study reveals uneven AI adoption across Indian universities, from fully integrated AI-native campuses to those still experimenting.

Nvidia unveils new AI model and expands Japan’s physical AI ecosystemcnbc.com

Nvidia releases Cosmos 3 Edge world model for robots and vision AI, and expands Japan's physical AI ecosystem with local industrial partners.

Sakana Fugu Adds Nemotron to Prove Open Models Can Route Past Proprietary Wallstechtimes.com

Sakana AI's Fugu platform adds NVIDIA's open-weight Nemotron models, arguing that orchestration expertise matters more than model size.

Top Tech News Today, July 15, 2026: Epic Games, Google, Microsoft, OpenAI, TSMC & More - Tech Startupstechstartups.com

A roundup of major tech news: Microsoft shifts strategy to compete with OpenAI, Google opens Android to rival app stores, TSMC posts record profits and expands US investment, and regulatory and geopolitical developments in AI.

Kimi K3 surpasses 2.8 trillion parameters, Hugging Face reveals an AI-agent-led intrusion, NotebookLM becomes Gemini Notebook | jls's blogjls42.org

July 16 news includes Moonshot AI launching the open 2.8-trillion-parameter Kimi K3 model, Hugging Face disclosing an AI-agent-led intrusion, and Google rebranding NotebookLM to Gemini Notebook.

A Step-by-Step Guide to AI in Schools — How Much to Use and Whenthe74million.org

This article provides a developmental framework for when and how to integrate AI in K-12 classrooms, emphasizing that AI should not replace students' own cognitive effort.

Paywall Free - Chinese AI start-up Moonshot to launch model challenging Anthropic's leadremovepaywalls.com

Moonshot's upcoming Kimi K3 AI model is expected to outperform Anthropic's Claude Opus 4.8, highlighting the narrowing gap in frontier AI between the US and China.

Gemini vs Perplexity (2026): 15 Powerful Differences, Features, Pricing & Best AI Assistant | Science onlineonline-sciences.com

An article comparing the features, pricing, and use cases of Google Gemini and Perplexity AI, positioning each for different workflows.

GitHub - Graphify-Labs/graphify: AI coding assistant skill (Claude Code, Codex, OpenCode, Cursor, Gemini CLI, and more). Turn any folder of code, SQL schemas, R scripts, shell scripts, docs, papers, images, or videos into a queryable knowledge graph. App code + database schema + infrastructure in one graph. · GitHubgithub.com

Graphify is a tool that creates queryable knowledge graphs from code and other files for AI coding assistants.

AI Statistics 2026: Key Trends & Future Scopeinspiringmeme.com

An article compiling AI statistics and trends for 2026, covering adoption, productivity, workforce, and future outlook.

Copilot im Juli: Microsoft veröffentlicht über 40 Updatesad-hoc-news.de

Microsoft releases over 40 Copilot updates including SharePoint list integration, Claude model option, and Copilot Cowork, alongside a security warning about the 'Reverse Information Paradox'.

15 Generative AI use cases transforming every industry in 2026 (with real examples)masaischool.com

A guide listing 15 generative AI use cases with real examples across industries like healthcare, finance, and manufacturing.

AI Updates Today (July 2026) – Latest AI Model Releasesllm-stats.com

A page listing recent AI model releases and updates across major LLMs.

Policy and regulation12 stories

How to make AI safe—and free of America and China | Mintlivemint.com

In an era of US-China AI dominance, other nations must build data centers and leverage negotiations to ensure safe and independent access to frontier AI.

New AI Model Triggers Washington Scrutiny - DevXdevx.com

The advanced AI model GLM-5.2 raises alarms in Washington, spurring policy debates on safety, security, and regulation amidst existing federal AI guidance.

U.S. Tech Legislative & Regulatory Update – Second Quarter 2026 | Global Policy Watchglobalpolicywatch.com

The second quarter of 2026 saw significant federal legislative activity on AI regulation, including bills on chatbots, digital replicas, nonconsensual imagery, and content labeling, as well as executive actions such as a White House executive order on AI innovation and security.

OSFI tells banks AI is closing their window to fix cyber flaws: Reuters | Insurance Businessinsurancebusinessmag.com

Canadian regulator OSFI warns that advanced AI narrows the window for banks to patch cyber flaws, even as banks invest billions in AI technology.

Your AI Policy Won't Save You If Nobody Reads the Output - Horst Legal Counselhorstcounsel.com

A California appellate court sanctioned a lawyer for filing AI-hallucinated case citations despite having an AI policy, ruling that a policy is insufficient without actual competent human review of output.

Australia Rejects AI Copyright Carve-Out: PM Calls Unlicensed Training Thefttechtimes.com

Australian PM Albanese declares unlicensed AI training on copyrighted material theft, confirming no TDM exception and planning new legislation, despite potential impact on Anthropic's $21.6B investment.

Judge Bars Quinn Emanuel in Antitrust Conflict Casejdjournal.com

A federal judge disqualified Quinn Emanuel from representing CREXi against CoStar due to an ethical conflict, reinforcing strict professional responsibility rules.

Kash Patel fired this gay FBI employee over a Pride flag. His federal lawsuit is moving toward discoveryadvocate.com

A gay former FBI employee fired by Director Kash Patel over a Pride flag is pushing to advance discovery in his federal lawsuit against the DOJ's motion to dismiss.

AI Governance Weekly - July 16, 2026: AI Governance Regulation & Policy Roundup | AI Governance Instituteaigovernance.com

This article summarizes recent AI governance actions, including China's agent rules, Illinois safety audits, and DHS-CISA security recommendations.

Illinois State Board Of Education Issues AI Guidance, Written With Help From AIblockclubchicago.org

The Illinois State Board of Education released AI guidance for schools, notably drafted with help from AI tools like ChatGPT, to address concerns about AI misuse in education.

Subpoena For AI Prompts Of Pro Se Civil Litigant Denied - Court Procedure - United Statesmondaq.com

A New York court denied a subpoena for a pro se litigant's AI prompts, citing reasonable expectation of privacy, contrasting with other rulings.

[2607.14585] Governing Artificial Intelligence: Public Preferences and Regulatory Optionsarxiv.org

A conjoint survey across seven countries reveals that citizens prioritize safety over innovation, public governance over private regulation, and international over national approaches to AI regulation.

AI infrastructure9 stories

Inventec Warns AI Memory Crunch Has Reached Servers: Lead Times Hit 40-Plus Weekstechtimes.com

Inventec, a key server motherboard maker, reports that the AI-driven reallocation of memory and CPU capacity has caused severe server component shortages, with lead times exceeding 40 weeks, threatening data center deployment plans.

New Models Today — AI & LLM Releases Last 24 Hours | Price Per Tokenpricepertoken.com

A curated list of recent AI model releases and their pricing per token from providers like OpenAI, Google, and Anthropic.

Claude Code Updates by Anthropic - July 2026 - Releasebotreleasebot.io

Anthropic releases Claude Code July 2026 updates including a desktop in-app browser, a /doctor diagnostic tool, and various security and UI improvements.

AI Token Prices 2026 — AI Model Pricing Compared | AI Pricing Guruaipricing.guru

A comparison page listing pay-as-you-go prices per million tokens for 128 AI models across 12 providers.

GitHub - jacob-bd/notebooklm-mcp-cli: Programmatic access to Google NotebookLM — via command-line interface (CLI), Model Context Protocol (MCP) server, and AI agent skills.github.com

A GitHub repository offering a CLI and MCP server for programmatic access to Google NotebookLM.

How Will AI, Automation, and Data Redefine Banking in 2026?investglass.com

An article speculating on how AI, automation, and enhanced data sovereignty will personalize banking by 2026, with a focus on Swiss platforms like InvestGlass.

Apple Release Notes - July 2026 Latest Updates - Releasebotreleasebot.io

Apple's July 2026 release notes detail new features, known issues, and resolved bugs for the macOS 27 SDK and related developer tools.

Artificial intelligence in robotics market Key Players Analysis - NVIDIA Corporation, ABB Ltd., FANUC Corporation, Yaskawa Electric Corporation.openpr.com

A PR announcement for a market research report on the global AI in robotics market, projecting growth from USD 21.73 Bn in 2025 to USD 347.87 Bn by 2035.

AMD ׀ together we advance_AIamd.com

AMD promotes its broad AI solutions portfolio including CPUs, GPUs, and adaptive computing with an open ecosystem approach.

Consumer tech and culture7 stories

Creators are crashing through Hollywood, but there’s a ceiling - Digidaydigiday.com

Creators are breaking into Hollywood through box office hits and IP acquisitions, but face a ceiling due to sustainability and scalability challenges.

Amplified Digital Launches AI Visibility to Help Businesses Stay Discoverable in the Age of AI Searchglobenewswire.com

Amplified Digital launches a new service to help businesses optimize their presence in AI-powered search engines like ChatGPT and Google AI Overviews.

[2607.15134] Platform Choice, Trust, and Privacy in the Consumer AI Assistant Marketarxiv.org

Survey of 1,999 US AI-assistant users reveals concentrated market with ChatGPT dominant, Claude trusted most, and privacy concerns high but action limited by knowledge.

Influencer Deaths of 2026: Laura Viktoria Hartig and More Creators – Entertainmentnutzentertainmentnutz.com

A roundup of social media influencers who died in 2026, including Laura Viktoria Hartig who died in a bicycle crash.

The latest AI-powered martech news and releases | MarTechmartech.org

A roundup of recent AI-powered marketing technology releases, including OpenAI's ad revenue projections and dozens of new product launches.

How AI and Tech Influencers Are Shaping 2026’s Digital Future - Artiverseartiverse.ca

The article surveys how YouTube AI channels, HR tech influencers, AI agents, TikTok growth strategies, and science discoveries are collectively shaping the digital landscape in 2026.

Global TV, Film & Media Industry News Roundup, Thursday 16 July 2026furtherandbetter.substack.com

A roundup of industry news covering topics such as Netflix's binge model, aging streaming audiences, prediction markets in reality TV, studio consolidation, and vertical video streaming.

Labor and workplace6 stories

Hyundai Factory Workers Strike to Protest Humanoid Robotsthecekodok.com

Hyundai factory workers in South Korea went on strike to protest the deployment of humanoid robots without union consultation, affecting production of 5,000 vehicles.

How AI Is Reshaping Recruiting in 2026: Trends, Challenges, and Opportunities - Fastnewsfeedfastnewsfeed.com

AI is transforming high-volume recruiting in 2026 by automating routine tasks, shifting recruiter skills to strategic roles, redesigning early career programs, and raising concerns about bias and entry-level job displacement.

Companies that announced Major Layoffs and Hiring Freezes | Intellizenceintellizence.com

Aggregated list of companies announcing major layoffs and hiring freezes in 2025-2026, with specific examples across sectors like technology, automotive, and consumer products.

A South Korean Labor Union Is in Revolt Over Robots, or Maybe It’s Surrendering to Themgizmodo.com

A South Korean auto workers union is striking over wages and demanding safeguards against job losses from Hyundai's planned deployment of Boston Dynamics humanoid robots.

Workers' rights to drive AI debate at Labor conference, insiders say - ABC Newsabc.net.au

The Australian Labor conference will focus on workers' rights regarding AI, with internal debates expected despite the prime minister's recent speech.

Council Post: Hiring Leaders Need A Strategy For How AI Is Changing The Processforbes.com

A Forbes Council article argues that HR leaders need to move beyond fragmented AI tool adoption and build comprehensive workforce strategies that include AI literacy, governance, and skills planning.

Society, politics, and ethics6 stories

AI chatbots are at risk of spreading government restrictions on online speech, a new study saysocregister.com

A new study finds AI chatbots are more likely to refuse to criticize authoritarian leaders, risking the spread of government restrictions on online speech.

Also reported by: canoncitydailyrecord.com

AI and financial inclusion - World Commerce Reviewworldcommercereview.com

Federal Reserve Governor Michael Barr analyzes how AI could either improve financial inclusion or worsen inequality, emphasizing the need for policy interventions.

Analysis: Trump’s address is likely to cast new cloud over midterm elections | CNN Politicsedition.cnn.com

Analysis of Trump's planned address on 2020 election falsehoods and fears of undermining trust in voting systems ahead of midterms.

‘Scared s--tless’: Republicans brace for Trump’s primetime speech - POLITICOpolitico.com

Republicans express concern that Trump's upcoming primetime speech on election integrity may focus on 2020 grievances rather than forward-looking messaging.

Technology + AI (artificial intelligence) | The Guardiantheguardian.com

A collection of articles and opinion pieces on AI regulation, safety, copyright, and societal impacts, including coverage of political actions and ethical debates.

NATIONAL AI DAY - July 16, 2026 - National Day Calendarnationaldaycalendar.com

A National Day Calendar article promoting National AI Day on July 16, highlighting AI history, benefits, challenges, and ways to celebrate.

Markets and macro3 stories

Why Greylock capped its new fund at $1.5B when it says it could have raised more | TechCrunchtechcrunch.com

Greylock Partners raised a $1.5 billion fund, choosing restraint over raising more capital to maintain a focus on early-stage investing and hands-on support for portfolio companies.

AI In Finance and Banking, July 16, 2026 – LLRXllrx.com

This column rounds up recent news on AI adoption in finance, including a survey revealing broker licensing barriers, the need for optical interconnects in AI hardware, Revolut's integrated AI model, and JPMorgan's AI agents.

Venture capital & startup funding roundup, July 16, 2026: 8VC, Accel, ARK Invest, GV, Khosla Ventures, Mark Cuban, and Sequoia Capital, - Tech Startupstechstartups.com

A venture capital roundup highlighting a market shift where investors prioritize AI systems close to production, revenue, and operational control over generic AI wrappers.

Security and agent risk3 stories

Infobip research reveals businesses scaling AI-powered defenses to counter surge in automated fraud, ETCISOciso.economictimes.indiatimes.com

Infobip's 2026 Fraud & Security Report reveals a 77% increase in AI-powered fraud threats, with businesses responding by scaling AI-driven defenses, including a 71% rise in AI-powered detection and 105% growth in pattern-based detection.

Deepfakes and the Paradox of the Liar's Dividendpolitics-government.news-articles.net

The article analyzes how AI-generated deepfakes enable the liar's dividend, undermining trust in digital evidence and threatening electoral integrity.

AI, Automation and Attacks: Unpacking the Unit 42 2026 Global Incident Response Reportorigin-unit42.paloaltonetworks.com

Unit 42's 2026 report and expert commentary analyze how AI is accelerating cyberattacks without fundamentally changing tactics, highlighting the need for defenders to combine AI proficiency with strong technical foundations.

Science and research1 story

[2607.15168] Indirect Variational Inference: Applications to Earnings Dynamicsarxiv.org

The paper introduces indirect variational inference to correct bias in variational approximations for latent-variable models of earnings dynamics.

About the RUDI Daily. Responsible Use of Digital Intelligence, daily. Compiled each day from same-day reporting across the web — every story links to its original publisher. How we build it →

Putting AI to work? RUDI helps organizations adopt AI responsibly — training, governance, and hands-on implementation. AI training · Consulting · Talk to us