AI news for 28 October 2026
Wednesday 28 October 2026 · Latentown
On Wednesday 28 October 2026, Latentown tracked 7 posts from 3 AI labs, 4 research papers and 45 news stories from 23 outlets. OpenAI was the busiest lab with 5 posts. The highest-ranked story was "Gemini 4 Carbon: Our next frontier model".
Top stories
- Gemini 4 Carbon: Our next frontier model
Gemini 4 Carbon is our strongest model with state-of-the-art cybersecurity and long horizon capabilities.
- OpenAI and ServiceNow expand partnership across enterprise workflows
OpenAI and ServiceNow expand their partnership to bring OpenAI models and agents into ServiceNow and ServiceNow workflows into ChatGPT and Codex.
- Legora is building the agentic operating system for legal work
See how Legora uses OpenAI models, ChatGPT, and Codex to accelerate legal workflows, expand contract analysis, and free lawyers to focus on higher-value work.
- Legora is building the agentic operating system for legal work
See how Legora uses OpenAI models, ChatGPT, and Codex to accelerate legal workflows, expand contract analysis, and free lawyers to focus on higher-value work.
- Google introduces cutting-edge Genie 4 world model
Google LLC today previewed Genie 4, an artificial intelligence model that can generate realistic three-dimensional environments.
From the AI labs
OpenAI
- OpenAI and ServiceNow expand partnership across enterprise workflows
OpenAI and ServiceNow expand their partnership to bring OpenAI models and agents into ServiceNow and ServiceNow workflows into ChatGPT and Codex.
- Legora is building the agentic operating system for legal work
See how Legora uses OpenAI models, ChatGPT, and Codex to accelerate legal workflows, expand contract analysis, and free lawyers to focus on higher-value work.
- Legora is building the agentic operating system for legal work
See how Legora uses OpenAI models, ChatGPT, and Codex to accelerate legal workflows, expand contract analysis, and free lawyers to focus on higher-value work.
- Evernote turns AI-driven discovery into paid subscriptions
Evernote used ChatGPT Ads to introduce its note-taking app to people working with AI. The team tracked how often free trials led to paid subscriptions.
- The future of the firm
How AI might enter the org chart—beneath, beside, and between us.
Google DeepMind
- Gemini 4 Carbon: Our next frontier model
Gemini 4 Carbon is our strongest model with state-of-the-art cybersecurity and long horizon capabilities.
Meta AI
- Meta Partners With Five US Universities to Accelerate Medical and Scientific Breakthroughs
Meta is bringing Superintelligence models, research funding, and direct researcher collaboration to US universities that are using our models to solve critical issues in science and medicine.
Research papers
- Intrinsic Distribution Matching
One-step and few-step text-to-image generation must accurately render the details specified in the prompt despite limited opportunities for refinement.
Earlier: DuoMatching: Joint-Marginal Distribution Matching for Few-Step Video Generation Town square · 1 October 2026
- Youtu-Parsing-Omni: One Encoder, One Schema, All Modalities
We present Youtu-Parsing-Omni, an omni-modal parser that converts documents, images, audio, and video into structured records grounded in space and time, providing a perception layer for agents that…
- Proving and Improving the Platonic Representation Hypothesis with JEPAs
Whether models trained on different modalities converge to the same representation of the world is central to AI.
- Does Scaling Toward AGI Depend on What System-1 Lets System-2 Execute?
Progress toward artificial general intelligence (AGI) has largely relied on scaling a resource-intensive executive, System-2, through larger models, more data and compute, longer reasoning, and more…
From the news wire
- Google introduces cutting-edge Genie 4 world model
Google LLC today previewed Genie 4, an artificial intelligence model that can generate realistic three-dimensional environments.
- VCs will make billions on Anthropic’s IPO, but Amazon is set to make much (much) more
Massive note conversions reveal the true extent of Amazon's ownership in Anthropic.
- Anthropic's IPO is risky business
Anthropic's much-anticipated IPO is giving potential investors some perspective on the risky business of artificial intelligence.
- Voice AI in vogue: Nvidia releases new open transcription model
Nvidia just released a free speech-to-text model that can transcribe up to 40 languages, but it only works on recordings, not live audio, so it’s a big step…
- The most important part of AI chatbots is being ignored
AI firms can increase the power of their chatbots without updating to new models.
- Polymathic AI and Princeton Researchers Release FlashMap, a Flow Map Drafter for 6.14x Faster LLM Decoding
FlashMap is a speculative decoding drafter from Polymathic AI, Princeton and USC ISI researchers.
- OpenAI launching election war room to track interference efforts
The company is bringing hundreds of employees together to respond to suspected efforts to use its tools to disrupt the midterms.
- OpenAI Plans to Buy Google’s Chips as Part of Its AI Build-Out
OpenAI intends to buy artificial intelligence computing chips from Alphabet Inc.’s Google, according to a person familiar with the matter, part of broader…
- xAI is buying out homeowners who promise not to sue over data center noise
A subsidiary of Elon Musk's xAI is offering to buy up homes in a Mississippi neighborhood, as long as residents don't join a lawsuit targeting the company's…
- Google alters its cloud playbook, enlisting partners to sell more AI chips
Google is opening its TPU business beyond its own cloud as custom chips become a bigger engine for growth.
- Meet Google DeepMind’s Genie 4: Real-Time World Model With Native Audio
Google DeepMind released Genie 4 as a research preview.
- Genius Group looks to pivot to data centers, partners with FingerMotion and BlueFlare for project in Canada
Edtech firm wants some of that sweet GPU cloud revenue
- Even after PHP2Shell, blast radius matters more than a CVSS score
PHP2Shell shows how quickly attackers can exploit critical vulnerabilities, often before organizations have time to patch.
- Should AI Decide Your Court Case? This Judge Says Yes
One of England’s top judges, Geoffrey Vos, believes that AI could adjudicate plenty of civil cases.
- Blockchain and crypto firm Blockmate Ventures enters neocloud market with purchase of Nvidia B200 cluster
Company has partnered with Scannit for the venture, with the latter to host the hardware in an undisclosed data center
- Nvidia's open Nemotron 3.5 Transcribe combines transcription, translation, and speaker recognition
Nvidia has released Nemotron 3.5 Transcribe, an open speech recognition model that handles transcription, speaker recognition, and translation and takes its…
- Nvidia releases all-in-one adaptable Nemotron 3.5 audio transcription and translation model
Nvidia Corp. today released a leaderboard-winning open artificial intelligence model for multilingual audio transcription and translation, along with an open…
- NVIDIA Nemotron 3.5 Transcribe and Granary v2: Intelligent Multilingual Transcription You Can Prompt and Adapt
- RoboParty Open-Sources Its RP1 Humanoid, From Mechanical Design and Joints to Control Software
RoboParty's RP1 is a 135 cm humanoid with 24 or 30 degrees of freedom, released as a full-stack open platform covering hardware, joint modules, firmware, robot…
- OpenAI teases Dots for Europe, but its Berlin DevDay was really about building a platform
At DevDay Exchange in Berlin, OpenAI mostly repeated what it had already announced, but it did share a few new details.
Earlier: OpenAI launches Dots that keep working without user prompts, plus other DevDay updates MediaNama · 1 October 2026
- Amazon tells engineers to stop ‘tokenmaxxing’ to look productive
Amazon has published new engineering guidelines warning developers and managers against ‘tokenmaxxing,’ arguing that AI usage metrics are ‘deeply flawed’…
- Why better data could be the key to AI-driven food R&D
As food companies navigate increasingly complex product development and safety requirements, connected data and generative AI could help scientists make…
- Intuit rewires accounting education as AI pushes new hires toward client work
As AI automates routine finance work, the repetitive tasks that once defined entry-level accounting are disappearing, and advisory services are moving to the…
- Meta’s Muse lands in Singapore with a guest list Amazon never got
When Meta switched on Muse in the US last month, it let its new personal AI agent wander into Amazon.com unannounced.
- How will AI shopping agents change Amazon and retail?
Amazon’s decision to block Meta’s Muse shopping agent looks, at first, like a technical dispute over access, security and authorisation.
- SoMark Open-Sources OXR-1.0, a 1.3B Document Parser That Leads Three Public Benchmarks
OXR-1.0 from Chinese document-AI team SoMark parses 13 element types into Markdown or JSON with layout coordinates, using a pluggable layout module and a…
- AI component hikes force the world's largest PC makers to return manufacturing capacity to China
PC manufacturing in China tends to be cheaper because of an established supply chain and workforce.
- KISS's Gene Simmons on AI in music industry: 'Learn how to use it, otherwise it will use you'
Gene Simmons of KISS fame is urging the music world to get on board with artificial intelligence.
- WorldOracle Tests Whether World Models Can Reliably Predict Real-World Executions
In one case showcased by WorldOracle, a robot's grasp missed, yet the world model predicted that the robot had successfully picked up the fruit.
- Unitree reports 70% surge in quarterly revenue, though AI investment weighs on net profit
Leading Chinese robotics company Unitree Robotics on Wednesday reported a 69.93 per cent year-on-year increase in quarterly revenue to 665 million yuan…
- Why the smartest AI models still need human context
- Macmillan Staff Say the Publisher Threw a Party for a Workplace AI Assistant That Was Later Killed
Workers tell WIRED that NOA—or “No Ordinary Assistant”—was unhelpful.
- AI Emergency Alerts Are Coming. Humans Might Not Be Ready
Human beings struggle with emergency communication.
- STAT+: As Epic’s dominance grows, its iron grip is choking innovation in health technology
A STAT investigation found that Epic selectively withholds access to patient data, disadvantaging rival software developers and stymying innovation.
- BNP Paribas CFO on Equities Trading, French Bonds, AI (Video)
- How Lao Gan Ma, China’s iconic chilli brand, got caught in an AI storm
For three decades, Lao Gan Ma’s glass jars, bearing the stern portrait of founder Tao Huabi, have been a fixture on Chinese dinner tables and a symbol of…
- Finding India’s Next AI Goldmine
A recent debate over investors’ lack of interest in India’s hardware and deeptech bets has brought some longstanding challenges in…
- Tsinghua Uni's Zhang on AI Innovation Outlook
Zhang Yaqin, Chair Professor and Founding Dean of AI Industry Research, and former President of Chinese tech firm, Baidu, discusses his outlook for AI…
- China Set to Lead in AI Models, Other Key Areas by 2030, BI Says
China is well placed to lead the world by 2030 in the large language models at the core of artificial intelligence, and will be at least competitive in every…
- ByteDance accelerates robotics commercialisation with new recruitment drive
ByteDance is accelerating its push into embodied artificial intelligence, stepping up recruitment for robotics talent as it seeks to move cutting-edge research…
- Inside Singapore’s AI push: Lighthouses, rice importers and robot labs
Singapore has a habit of announcing its future in instalments.
- Huawei's VitaAAU Brings RetinaMIMO to Upper 6GHz With 256 Channels and Over 1,500 Antenna Elements
At Decoding China 2026 in Shanghai, Huawei showed VitaAAU, an Upper 6GHz active antenna unit built for uplink-heavy AI traffic.
- AgiBot's Genie Master Lets G2 Robots Diagnose Their Own Failures on a Real Convenience-Store Shift
AgiBot's new embodied agent framework links an AI agent to G2 robots through a four-part execution protocol and two self-improvement loops.
- The most dangerous thing about AI isn’t hallucination. It’s that it stopped showing its work
Everyone is worried about the wrong AI failure.
- How to fund your robotics company: not another fundraise
Every story links to its original source. Ranking is by automatic rules, not an editor. Something wrong? Tell us.