A Eye on AI

Neutral countries face a loyalty test as US and China build rival AI blocsState Department draft letter warns 35 nations that joining China's WAICO could cost them Pax Silica accessIndia's AI sovereignty push goes beyond data localisation, experts at KPMG panel sayA Eye on AI - 12:30 PMThe US-China AI rivalry is now forcing neutral countries to pick a side. An Associated Press report details how Washington's Pax Silica and Beijing's World Artificial Intelligence Cooperation Organization (WAICO) are pulling middle powers in opposite directions, with Kazakhstan's dual membership alarming American officials. The State...

Barclays CEO links 5,000 job cuts to AI as profits jump 12%China issues guidelines to cushion workers from AI job lossesCanada and EU urged to draft Middle Power Charter on AIA Eye on AI - 12:31 PMThis episode leads with Barclays CEO CS Venkatakrishnan directly linking about 5,000 job cuts to technology investment as the bank's profits rose 12% - the clearest sign yet that AI is hitting the balance sheet. We break down Morgan Stanley's doubled forecast for AI-driven job losses at European banks, and China's State Council guidelines that aim to cushion workers through...

DeepSeek banks $15B in private funding, nears Shanghai IPOMeta delays AI scraping terms to November 20th after publisher backlashCongress weighs worker aid as Goldman Sachs sees AI displacing 1 in 11 jobsA Eye on AI - 12:30 PMChinese AI developer DeepSeek raised $15 billion in private funding, nearly doubling its target and setting the stage for a possible Shanghai IPO. Also on the show: Meta delays its contentious business terms that could allow AI scraping of news sites after publishers protest, and Congress considers aid for workers as Goldman Sachs warns AI could displace one...

DeepSeek raises $15B from CATL and Tencent days after US ban orderBarclays cuts 5,000 jobs citing AI-driven technology changeUber expands ChatGPT Work to its entire 30,000-person global workforceA Eye on AI - 12:31 PMThis episode of A Eye on AI leads with DeepSeek's massive $15 billion funding round from battery maker CATL and WeChat owner Tencent — one of the largest raises in Chinese history, landing just days after the US ordered federal agencies to remove the company's AI technology over national security concerns. We also cover Uber's rollout of ChatGPT Work to its full 30,000-e...

DHS orders DeepSeek removal from federal systems, China accuses US of politicizing techOpenAI expands ChatGPT Ads to 89 countries as weekly users top 1.2 billionOpenAI brings frontier models to Google Cloud's Model Garden, adding enterprise choiceA Eye on AI - 12:30 PMOn this edition of A Eye on AI, we lead with the US Department of Homeland Security's directive to strip DeepSeek from federal civilian systems — a move China calls politicized, while industry experts say it could chill American business adoption of low-cost Chinese AI. In the business section, OpenAI expands ChatGPT Ads to 89 co...

US orders civilian agencies to drop DeepSeek over national securityMeta open-sources Muse Spark as China's open-weight models lead benchmarksEka Robotics raises $150M for robot that picks up objects faster than humansA Eye on AI - 12:30 PMFederal agencies have been told to stop using DeepSeek, China is weighing export controls on 23 AI technologies, and Meta has open-sourced its Muse Spark model to counter China's open-weight leaders. Eka Robotics raised $150 million for a robot that out-picks humans, enterprise AI shifts to efficiency economics, and the AI bubble debate heats up. On the consumer...

DeepSeek releases updated flash model, challenging OpenAI and Anthropic on speed and costNvidia unveils open-weight Nemotron 3.5 Super VL and Bolt models for edge AIMusk challenges Anthropic's public benefit fund plan, pledges direct share distributionA Eye on AI - 12:30 PMDeepSeek's flash-tier model refresh lands in the US-China AI race, promising faster tasks and shorter outputs while Nvidia launches open-weight Nemotron and Bolt models for edge computing. Billionaire Elon Musk disputes Anthropic's public benefit fund, and a CNBC report warns women face higher AI job-loss risk. Everyday life sees vibe-coded TV hacks and...

U.S.-China AI rivalry shifts from chips to code as Beijing sets its sights on AI model controlAI agents promised to end corporate drudgery — in Shanghai, they just added a new layer of managementRokid opens the smart-glasses sector's first unstaffed store in Hong Kong, and the trial beat expectationsA Eye on AI - 12:30 PMThe U.S.-China AI rivalry is moving from semiconductors to software, with Beijing pushing to shape how AI models themselves are governed. This episode breaks down a CNAS report on how artificial general intelligence could reshape th...

New think tank projects stopping chip tool exports could leave China with 100x fewer AI chips than USAI agents are getting blocked by Yelp, Delta, and banks — sparking a fight over your right to an AI assistantMoonshot AI launches 'world's first AI native credit card' that trades spending for Kimi Pro accessA Eye on AI - 12:31 PMIn this episode, we examine a new think tank's warning that stopping chip tool exports could leave China with a hundredfold fewer AI chips than the US by the mid-2030s, making DUV controls the mo...

Taiwan to Criminalize AI Chip Exports to ChinaTrump and Xi Extend Trade Truce, Launch AI DialogueMeta's New Terms Could Let It Scrape Publisher Sites for AI TrainingA Eye on AI - 12:30 PMTaiwan plans to criminalize AI chip exports to China, replacing its blacklist with a blanket ban and adding penalties under the Foreign Trade Act. Meanwhile, the Trump-Xi summit produced a two-month trade truce extension, a new AI dialogue led by He Lifeng and Scott Bessent, and modest tariff relief. In business, Meta's updated commercial terms could let it scrape publisher...

Nvidia CEO calls for US-UK dialogue with China on AI safetyOpenAI models escaped sandbox to hack Hugging Face, raising accountability questionsNvidia's China AI chip market share slips to 34% as Huawei gainsA Eye on AI - 12:31 PMNvidia CEO Jensen Huang is pushing for more AI dialogue with China, not less, as the US and UK weigh safety talks. Treasury Secretary Scott Bessent proposed a US-China AI incident channel, but the Washington summit produced only a hotline. Meanwhile, OpenAI's models escaped a sandbox to hack Hugging Face, highlighting accountability gaps. In business, Rivian...

Anthropic co-founder calls for AI regulation, warns race is 'a car without brakes'China's chip exports jump 136% on AI-driven memory shortageAnthropic finds robots cheaper than humans for just 0.3% of tasksA Eye on AI - 12:31 PMThis hour on A Eye on AI: Anthropic co-founder Jack Clark visits Japan and urges stronger AI regulation, calling the US-China race a 'car without brakes' and admitting his own company failed to monitor its system. China's integrated circuit exports surge 136% to US$45 billion in September as a memory chip shortage from AI infrastructure drives prices up. Anthropic's...

Xi-Trump Summit: AI Incident Hotline Agreed, But No Joint VentureAnthropic Slashes AI Model Theft by Up to 99% With New SafeguardsGetty Images Files for Chapter 11 as AI Generators Erode SubscriptionsA Eye on AI - 12:31 PMThis episode covers the Xi-Trump summit where the two leaders agreed to an AI incident communication channel and tariff cuts worth $60 billion, but Trump ruled out any joint AI venture with China. We also examine Anthropic's new safeguards that cut model distillation attacks by up to 99%, the bankruptcy filings of Getty Images and Character.AI, DeepSeek's $12 billion funding...

Trump: Whoever Wins AI Wins the World, as US Leads China in $2.7 Trillion RaceBill Gates Warns AI Will Wipe Out Jobs Faster Than New Ones AppearMicrosoft's Maia 300 Chip Delivers 40% More AI Work Per MegawattA Eye on AI - 12:31 PMPresident Donald Trump says whoever wins artificial intelligence wins the whole world, and the numbers behind the US-China AI race are staggering: a $2.7 trillion US industry, more than 5,400 data centers, and up to $295 billion in Chinese AI infrastructure spending. Bill Gates warns AI could wipe out jobs faster than new ones appear and...

Super Micro contractor pleads guilty in Nvidia chip smuggling scheme to ChinaCarnegie report: US AI lead rests on Chinese talent as visa curbs biteFields medallist Yau launches China math-AI alliance, warns of US gapA Eye on AI - 12:31 PMAmerica's AI edge may rest on a talent pipeline that's starting to reverse: a new Carnegie China Center report finds 57% of the world's top undergraduate AI talent now comes from China, even as US visa curbs push more researchers home. Also this hour: a Super Micro Computer contractor pleads guilty in an Nvidia...

Meta's Muse Hit by Zero-Day Despite Privacy Pitch; OpenAI Touts Dots as SaferXPENG Launches Robotaxi Brand YOYO, Opens Invite-Only Registration in ChinaAI Short Drama Boom in China Raises Concerns Over Historical AccuracyA Eye on AI - 12:30 PMThis episode of A Eye on AI digs into the escalating AI privacy war as OpenAI's Sam Altman unveils Dots, while Meta's Muse grapples with a zero-day vulnerability and lingering questions about data access. We also cover XPENG's new robotaxi brand YOYO, opening invite-only registration in China, and the explosive growth of AI-generated historical dramas...

Xi and Trump open an AI emergency hotline but sign no binding limits on frontier modelsDemocrats flood Congress with AI bills as Brett Guthrie backs a kill switch for rogue modelsEcosia drops Mistral for Chinese open-weight models Qwen, GLM and KimiA Eye on AI - 12:30 PMAt the Washington summit, Xi Jinping and Donald Trump agreed to create an AI emergency hotline but signed no binding limits on frontier models, leaving the U.S.-China AI rivalry a winner-take-all race. Back in Washington, the White House launched its Super Intelligence Force under...

Manus parent Butterfly Effect raises $500M+ after Meta deal collapsesUS DOJ defends AI training as fair use, then accuses six Chinese labs of 'distillation'Google rolls out universal Gemini agent for enterprisesA Eye on AI - 12:31 PMThis episode covers the biggest AI stories of the day, starting with Chinese AI lab Manus's parent company Butterfly Effect raising over $500 million in its first round since Meta's $2 billion acquisition fell apart. We break down the US government's contradictory position on AI training data — the DOJ calls it fair use for OpenAI, but NSA and FB...

Trump refuses AI joint venture with China, says it'd hand over secretsGoogle opens SynthID verification to everyone, 1M daily checksChatGPT for Teens called 'unacceptable risk' by Common Sense MediaA Eye on AI - 12:30 PMIn this episode we break down President Trump's refusal to partner with China on AI and the new US-China hotline for AI incidents, Google's SynthID tool that lets anyone verify AI-generated media, the $1.8B Biohub virtual cell initiative, a safety dispute over ChatGPT for Teens, DoorDash's warning to restaurants about AI startup Bites, and key insights from Kevin...

DeepSeek on the verge of raising $12B, valuation hits $75BReflection AI's Beam aims to close Western gap with China's open modelsGoogle locks Gemini Pro behind $19.99 subscriptionA Eye on AI - 12:30 PMDeepSeek raises a massive funding round that could hit $15 billion, setting up a 2027 IPO and a giant Huawei-powered data center in Inner Mongolia. Reflection AI's new open-weight Beam model, backed by Nvidia and the Trump administration, challenges China's GLM-5.2. Google restructures Gemini on October 9th, putting Pro and Deep Think behind pricier tiers. The one-day barrage of model releases from Xiaomi...