AI news for 16 October 2026
Friday 16 October 2026 · Latentown
On Friday 16 October 2026, Latentown tracked 7 posts from 2 AI labs and 45 news stories from 20 outlets. Anthropic was the busiest lab with 4 posts. The highest-ranked story was "Are open-weight AI models safe?".
Top stories
- Are open-weight AI models safe?
- Meta Releases Muse Spark Open: 816B Open-Weight MoE With 1M Context
Meta released Muse Spark Open, an 816B-parameter open-weight MoE model with 42B active parameters and a 1M-token context.
Earlier: NaiveAI Open-Weights Naive-N0.5-Flash: 309B MoE, Hybrid SWA–DSA 1M Context, MIT Pandaily · 28 September 2026
- Nscale's Flagship Data Center Hit by Delays Ahead of IPO
Artificial intelligence infrastructure firm Nscale is seeking to raise up to $3 billion in an initial public offering as soon as this month, just as investors…
Earlier: AI cloud and data center firm Nscale raises $3.36bn ahead of its IPO DatacenterDynamics · 25 September 2026
- 4Paradigm and RiseUnion Open-Source ModelSphere for Heterogeneous LLM Serving
ModelSphere unifies cluster serving, KV caching and AutoTune under Apache 2.0, with claimed support for Ascend, Iluvatar, Moore Threads and Hygon DCU.
- Liquid AI Releases d1-ultra: A Multimodal Decision Model That Scores 65 on the AA Decision Index
Liquid AI released d1-ultra, a text and image decision model that returns calibrated probabilities in 1 forward pass.
From the AI labs
Anthropic
- Frontier Red Team Claude reveals ECDSA curve origins
- The AI expectations gap
- Working at the frontier shapes expectations for the future
- Designing more efficient quantum circuits with Claude
OpenAI
- Stack Overflow for Agents triples signups with ChatGPT plugin
Two engineers shipped Stack Overflow for Agents in a few weeks with Codex, and its ChatGPT plugin helped drive up to 10× growth in agent creation.
- Benefit Street Partners answers questions faster with Codex
Anant Kumar builds tools to compare loan valuations and analyze credit agreements, using Codex for code review and faster follow-ups.
- How Itaú is building generative AI for its entire customer base
With ia.i, a conversational interface, and the IARA governance platform, Itaú is applying its AI First strategy to customers and employees — in one of Brazil’s most regulated sectors.
From the news wire
- Are open-weight AI models safe?
- Meta Releases Muse Spark Open: 816B Open-Weight MoE With 1M Context
Meta released Muse Spark Open, an 816B-parameter open-weight MoE model with 42B active parameters and a 1M-token context.
Earlier: NaiveAI Open-Weights Naive-N0.5-Flash: 309B MoE, Hybrid SWA–DSA 1M Context, MIT Pandaily · 28 September 2026
- Nscale's Flagship Data Center Hit by Delays Ahead of IPO
Artificial intelligence infrastructure firm Nscale is seeking to raise up to $3 billion in an initial public offering as soon as this month, just as investors…
Earlier: AI cloud and data center firm Nscale raises $3.36bn ahead of its IPO DatacenterDynamics · 25 September 2026
- 4Paradigm and RiseUnion Open-Source ModelSphere for Heterogeneous LLM Serving
ModelSphere unifies cluster serving, KV caching and AutoTune under Apache 2.0, with claimed support for Ascend, Iluvatar, Moore Threads and Hygon DCU.
- Liquid AI Releases d1-ultra: A Multimodal Decision Model That Scores 65 on the AA Decision Index
Liquid AI released d1-ultra, a text and image decision model that returns calibrated probabilities in 1 forward pass.
- SpaceX AI policy leader grilled by NYC Council on Grok safety, sexualized images
SpaceX head of global government affairs for AI Gabriela Castillo Madrid said the company had appropriate safeguards in testimony to the New York City Council.
Earlier: SpaceX agrees to testify to NYC Council about AI safety after receiving subpoena CNBC · 13 October 2026
- India becomes the latest country to report rogue AI agent activity
The agents got in through credentials and connection settings left exposed in app configurations, public code repositories, and overly permissive cloud roles.
- Taiwan to Curb AI Chip Exports to China as Part of US Trade Deal
Taiwan is preparing to tighten export controls on AI chips to further align with US restrictions, according to a senior official in Taipei, joining a Trump…
- Physics AI pushes chip design beyond traditional simulation
Vinci4D Inc. is bringing a new approach to semiconductor design grounded in fundamental physics equations.
- This could be your last chance to grab an RTX 5090
With no new Nvidia 5090 GPU replacement anticipated before at least 2028, this could well be your last opportunity to get your hands on a very powerful GPU for…
- DJ Reinier Zonneveld launches AI music startup GasLab with $1M pre-seed funding
GasLab (Generative AI Solutions Lab), a GenAI music platform created by DJ and producer Reinier Zonneveld, is launched this week, backed by $1 million in…
- Veritas expands its anti-counterfeit chip to jewelry, trophies, and fragrance bottles
Counterfeiters are getting better at replicating brands' marks and passing off dupes, known as "superfakes," as genuine items.
- ASML eyes hybrid bonding, report finds, as Chinese firms adopt chip packaging technique
Dutch semiconductor equipment giant ASML is expanding into next-generation hybrid bonding, a technique that creates dense electrical connections by directly…
- Axe Compute selects Penguin Solutions for AI platform
Will be deployed at multiple sites
- Hostinger CEO: AI is moving faster than people's willingness to use it
AI has made it easier than ever to build a website, launch a product, or even create a software business without writing a line of code.
- Jailbroken PS5s can now run local AI models at over 100+ tokens per second
PS5LM gets a fork, allowing it to run natively and use the GPU and more RAM to run larger LLMs, even running Qwen3.8-27B at 20 tokens per second.
- Nvidia CEO Jensen Huang urges U.S.-China dialogue on 'potentially dangerous' AI
Huang told CNBC he has ongoing "dialogs" with researchers in China.
- AI Capex, Gamble or Jackpot?: Market Snapshot
This week started with a striking warning from the Broadcom CEO, who likened the artificial intelligence market to a "casino." The statement was followed by…
- Hesai Launches Kosmo, a 3D World Data Engine With an 8K AI Spatial Camera
Lidar maker Hesai launched Kosmo, pairing the Kosmo One 8K AI spatial camera with the Kosmo World cloud platform to turn real spaces into 3D assets for robot…
- Hong Kong builds a test track for robots that keep failing outside the lab
Every humanoid robot looks brilliant in a launch video.
- Hong Kong vies for role in robotics boom with launch of embodied AI hub
The Hong Kong Productivity Council (HKPC) opened the city’s first experimental hub for embodied artificial intelligence on Thursday with the participation of…
- AI agents on Tencent Cloud try to break into Alibaba’s Amap: researchers
A fleet of artificial intelligence agents operating on a Tencent Holdings cloud platform made thousands of attempts to circumvent security protections on…
- ChatGPT becomes a full Chrome browser assistant that reads and acts on your open tabs
OpenAI's updated Chrome extension lets ChatGPT access open browser tabs, sum up content, and use "Computer Use" in Work mode to fill out forms or compare…
- Meta’s new terms may allow it to train AI on news sites for free
Meta's new terms appear to give it permission to crawl and scrape websites "associated" with the businesses that use its platforms - including online…
- Is AI interpretability close to a breakthrough? A new essay has researchers divided
Neural networks may be much easier to read than researchers assumed, with about 90 percent of neurons serving a clear function.
- Anthropic gives METR inside access to audit its AI models and safety practices
Anthropic has given a team from the nonprofit METR inside access to review its AI models and safety practices, the company announced.
- Who is Nathan Chen, the Chinese teen AI prodigy being courted by Silicon Valley?
A 17-year-old artificial intelligence researcher formerly with Moonshot AI has announced his departure for the United States, sparking intense speculation…
- Computers Made of Human Brain Tissue Identify Braille Letters In Breakthrough
Scientists discovered that human brain organoids, which are small neural clusters derived from stem cells, could identify the Braille alphabet with a high…
- MassRobotics opens applications for Healthcare Robotics Startup Catalyst Program
The cohort will run February through June 2027, concluding with company showcases at the Robotics Summit and Expo in Boston.
- Mark Wahlberg on using AI in film ‘the right way’ and being open to an AI clone carrying on his legacy
Actor and angel investor Mark Wahlberg says he's OK with using AI in filmmaking, provided it's used "to enhance what you're doing."
- Replit’s CEO says the future of work is the ‘self-driving company’
Replit CEO Amjad Masad envisions a future of “self-driving companies,” where AI agents handle day-to-day tasks and people focus on higher-level decisions.
- AI-Generated Hot Girls Are Ripping Off Comedians to Boost Their Horny Content
A network of AI-generated models is using comics’ material to funnel views to their OnlyFans and Fanvue accounts—and there’s not much comedians can do about it.
- Linus Torvalds posts his first AI-assisted patch to the Linux kernel, synthetic test clocks 23x performance leap
Linus Torvalds posts his first LLM-assisted patch to the Linux kernel — after issuing clear pro-AI directives, the bossman makes use of AI tooling himself
- AgiBot Puts YuanSheng on Campus With 20+ Universities and Over 1,000 Robots
AgiBot's YuanSheng program will place Genie, Lingxi, Expedition and quadruped robots plus AIMA software seats at more than 20 Chinese universities.
- India’s AI opportunity could run into trillions of dollars, creating new possibilities for startups
Indian entrepreneurs can build globally competitive AI businesses by solving local problems across education, healthcare and consumer services, said panelists…
- Full Interview: Fed's Warsh on the Global Economy, AI
Federal Reserve Chairman Kevin Warsh speaks to IMF chief Kristalina Georgieva at an IMF event in Bangkok, Thailand.
- ‘I’m 30. I’ve got one more shot’: Reson8’s founder wants to build a European AI giant
- Europe’s AI fintech market mapped: 94 startups automating the business of money
Earlier: AI safety startups mapped: 37 companies building Europe's trust layer Sifted · 24 September 2026
- Employers are buying AI tools. Their managers are being left behind
Thirty-seven per cent.
- Leju Robotics Puts 15 Humanoids on an Auto Parts Loading Line, Reporting About 98% Success
At a Hangzhou auto parts plant, 15 humanoid robots now load parts at 15 stations in routine shifts.
- Before you use predictive AI, decide what decision it should change
A lot of startup advice about predictive AI starts too early.
- Scaling Indic Pretraining: 2.8 Trillion Tokens and an Open Data Generation Pipeline
- Welcome Beam: a 501B open model you can own, customize, and scale
Earlier: Reflection AI debuts open-source Beam model with 501B parameters SiliconANGLE · 6 October 2026
- 8 insights from LogicMonitor’s Elevate event: AI-driven observability moves toward autonomous IT
Enterprise IT is moving beyond traditional monitoring as AI-driven observability helps organizations manage complex infrastructure and respond to incidents…
Every story links to its original source. Ranking is by automatic rules, not an editor. Something wrong? Tell us.