Models, products and AI startups, near-duplicates collapsed to the most credible source
Union Square Ventures Doubles Fund Size, Shrinks Team for AI EraNew York-based venture firm Union Square Ventures is nearly doubling its core fund size and reducing its general partnership to just four investors to better compete in the AI race.
Gudea Lands $7M Seed Funding to Predict Viral Online NarrativesGudea raised $7M in seed funding co-led by Mudita Venture Partners and Silicon Road Ventures, bringing total funding to about $15M. The startup uses AI to predict narrative spread across 600+ social platforms, with revenue up 357% YoY and customers including major AI firms and telecom, entertainment, and retail businesses.
audio.cpp update cuts Higgs Audio TTS VRAM use by 48%audio.cpp landed runtime optimizations: Higgs Audio TTS now runs in about 6GB VRAM, a 48% peak-memory cut; HTDemucs is 2.21x faster on GPU and PocketTTS 2.23x faster on CPU, with several TTS models seeing 6-21% VRAM reductions. The project now supports 110+ audio model families and 190+ variants, and adds an experimental WebUI generation history feature.
JPMorgan's Gehl on Tech & AI Fundraising, Capex, IPOMatthew Gehl, Co-head of International Technology Investment Banking at JPMorgan, spoke with Bloomberg's Tom Mackenzie at the 2026 JPMorgan Tech Stars event in London on Oct. 6, discussing tech and AI fundraising, capex, and IPOs.
LegalOn cuts daily Codex costs by 65%LegalOn cut estimated daily Codex costs by 65% while maintaining development speed by matching Astra, Sol and Luna to tasks and managing budgets strategically.
Pollo AI turns ideas into ads with OpenAIPollo AI uses GPT-5.6, GPT-6 Astra, and GPT-Image-2.5 to help creators turn bold ideas into detailed images and cinematic video ads.
Vesta raises $30M to bring swarms of AI agents to mortgage lendersVesta, an AI-native mortgage origination software startup, announced a $30 million round led by Conversion Capital, with participation from Citi Ventures and Andreessen Horowitz. Its AI agents automate loan origination; revenue grew 12x year over year, it helps originate over $100 billion in loans annually, and it has raised $85 million to date, though its market share remains under 5%.
Broadcom in talks for over $50B financing for OpenAI chipsBroadcom is in talks with lenders for over $50 billion in financing for custom AI chips it is developing with OpenAI, after a $60 billion financing for Anthropic. Broadcom expects about $230 billion in AI chip revenue in fiscal 2028.
Nvidia bets on physical AI with Halos safety systemNvidia is investing billions in physical AI such as robotics and self-driving cars and has built a full-stack safety system, Halos, so companies can reduce the risk of machines harming people. Halos for Robotics launched in June 2026.
Europe venture funding hits $25B in Q3, a 4-year highCrunchbase data shows European startup funding reached $25 billion in Q3, up 77% year over year and the strongest quarter in four years, with AI taking 75%. Mistral's $3.5 billion round was Europe's largest ever.
IBM's Rossi: Trust Is the Real AI ChallengeIBM's Francesca Rossi says security and competitiveness are not opposing goals, and that the real edge will come from trustworthy systems rather than the most powerful models, warning some AI risks are underestimated. The interview follows Dario Amodei's call for a slowdown and Pope Leo XIV's remarks.
Poll: 64% of Americans Say AI Develops Too FastA new AP-NORC poll finds 64% of Americans say AI is developing too fast, versus only 8% who say too slow. Most want government to keep AI under human control and protect workers, yet few trust either party on the issue, and Trump's approval is low.
OpenAI Hires Second SpaceX Sales Exec in a MonthOpenAI appointed Ismail Elmas as VP of sales for Europe, the Middle East and Africa, its second sales hire from Elon Musk's SpaceX since September.
South Korean banks hacked, possibly by one person using AI toolsA CrowdStrike report says several major South Korean banks were hit by a cyberattack last week, possibly carried out by one person using the open-source penetration tool ARTEX plus AI models DeepSeek v4.1-Flash, GLM-5.3, Grok 4.6 and Claude Code.
Analyst Calls Anthropic IPO 'Most Ridiculous' of 2026, Values Firm at $150BIndependent research firm New Constructs labeled Anthropic's potential IPO as the 'most ridiculous IPO of 2026,' valuing the company at just $150 billion versus a possible $2 trillion market cap. It argues Anthropic lacks a viable business and the IPO would mainly create liquidity for Wall Street backers.
Teen using Claude for hike route planning rescued by helicopterA 16-year-old planned his route on Crown Mountain near Vancouver using Anthropic's Claude and had to be airlifted off a steep rock face that requires climbing gear. He doesn't blame the AI but says he'll never use it for route planning again.
Saluki 27B claims 96% of Qwen 3.8 performance at 1/7 sizeA LocalLLaMA user posted about the Saluki 27B model, claiming it reaches about 96% of Qwen 3.8's performance at roughly one-seventh the size, and asked the community for feedback.
Anthropic, Google, OpenAI Agree to UK Privacy Changes, Regulator ConfirmsAnthropic, Google, and OpenAI have agreed to UK privacy changes, as confirmed by the regulator. Specific details of the changes were not disclosed in the report.
Philly tenants fight to limit AI property managementWHYY reports that Philadelphia tenants are pushing to limit AI use in property management, citing concerns about its impact on renters' rights.
Tristan Harris' nonprofit lays off most staffThe Center for Humane Technology is cutting nine staffers, including executive director Julie Guirado and senior director of strategy Camille Carlton, moving to a founder-led model. Its board said splitting between advocacy and think-tank work stretched it thin, so it will focus on public education, targeted briefings, and private coordination, dropping litigation support and AI tool evaluation.
Turing Institute Head: Frontier AI Models 'Extraordinarily Expensive'Alan Turing Institute head George Williamson urges the UK to prioritize cost-effective application-specific AI models for public services over the most advanced products. Research shows tailored models approach frontier performance at much lower cost, and he notes ordinary people feel excluded from AI discussions.
Japan cyber threats rise as AI lowers barrier for hackersThe Japan Times reports rising corporate data breaches in Japan, with experts warning AI is lowering the barrier for hackers and scammers, as new laws allow more preemptive measures.
Report suggests OpenAI has clawed back market share from Anthropic in 2026A report suggests that OpenAI has regained significant market share from rival Anthropic in 2026.
MIT: Humanoid robot breakthroughs won't change life soonDespite Musk calling Optimus Tesla's biggest product ever with a possible 2027 launch, Huang saying humanoids will match human ability this year, and Morgan Stanley projecting nearly 1 billion units and a $5T market by 2050, today's robots still fail often, so breakthroughs won't change daily life soon.
OpenAI claims AGI era has arrived — has GPT-6 Astra really shown general intelligence?OpenAI claims we have entered the AGI era, pointing to its new GPT-6 Astra, but the article questions whether the model truly demonstrates general intelligence. If true, the claim would mark a major milestone in AI, yet verifiable evidence is still lacking.
JPMorgan Private Bank says AI set to drive strong earningsNataliia Lipikhina, head of EMEA equity strategy at JPMorgan Private Bank, says AI continues to underpin global equity markets, particularly tech. She notes strong demand and constrained supply, and that the AI story continues.
AI tool adoption nearly doubles participation in WorldQuant quant contestWorldQuant said about 156,000 people signed up for its International Quant Championship this year, nearly double last year's roughly 80,000, as wider AI tool access lowered barriers to building trading models. Victor Ayebameru, a student from Nigeria, took first place; over 40,000 women participated and about 28% of entrants came from Africa.
Autonomous industrial AI needs new governance framework for safetyIndustrial AI is entering a new phase, with foundation models, physical AI, and agentic AI enabling automation but raising safety risks. AVEVA CTO Arti Garg emphasizes responsible deployment, arguing AI should augment rather than replace human decision-makers, with governance frameworks. Data integration and sustainability are also key.
AI key updates and developments (Sep 24-Oct 7)This roundup covers recent legislative and regulatory developments on AI and data protection in Malaysia, Greece, the EU and Minnesota, USA.
Anthropic's Claude AI adds Chinese-language optionsAnthropic has added Chinese-language options to its Claude AI, improving accessibility for Chinese-speaking users.
The A.I. Daily