Big week in AI and automation: Rishi Sunak jumped into advisory roles at Microsoft and Anthropic, OpenAI’s Sora topped ChatGPT’s iOS debut, Firefox baked in Perplexity, and Google rolled out Gemini’s “Help Me Schedule.” On the research front, a new study found that how people treat service robots can influence bystanders’ behavior—design cues and ethics guidelines matter.
Policy moved too, with California approving rules for AI companion chatbots, including age checks, safeguards, and disclosures. On the calendar, IROS in Hangzhou and the BGI Summit in Istanbul/online are around the corner. And in capital, JPMorgan announced a $1.5T push into strategic and frontier tech, while Baselane, PEAK:AIO, Phaidra, and Heidi raised fresh rounds—and Radical Ventures closed a $650M AI fund.
Industry & Enterprise
Former UK Prime Minister Rishi Sunak Takes Senior Advisory Roles at Microsoft and Anthropic
Rishi Sunak, U.K. prime minister from 2022–24, has taken senior advisory roles at Microsoft and Anthropic, prompting Acoba to warn that his access to privileged government information could give Microsoft an unfair edge amid its U.K. contracts and a £2.5 billion data-center investment announced during his tenure. Sunak said he won’t advise on U.K. policy and will donate his salary to the Richmond Project; he is also a senior adviser to Goldman Sachs. (AI Insider)
OpenAI’s Sora Surpasses ChatGPT’s iOS Debut with Record-Breaking Launch
OpenAI’s Sora iOS app logged 627,000 U.S. downloads in its first week—surpassing ChatGPT’s 606,000 debut—and quickly hit the App Store’s top spot after its Sept. 30, 2025 launch, per Appfigures. Head of Sora Bill Peebles said the invite-only app crossed 1 million downloads in under five days, despite being limited to iOS in the U.S. and Canada and powered by the Sora 2 video model. (AI Insider)
Thinking Machines Lab Co-Founder Andrew Tulloch Departs to Join Meta as AI Talent Wars Escalate
CoreWeave (Nasdaq: CRWV) agreed to acquire Monolith AI to fuse its AI cloud with Monolith’s simulation and test-driven ML, aiming to shorten industrial R&D cycles using anomaly detection, test-plan optimization, and next-test recommendations already deployed at Nissan, BMW, and Honeywell, with McKinsey estimating 20–80% efficiency gains. Terms weren’t disclosed and closing awaits customary conditions; the deal complements CoreWeave’s OpenPipe and Weights & Biases moves and advances its bid to be industry’s default AI cloud, underscored by its Aston Martin F1 partnership. (AI Insider)
Mozilla Integrates Perplexity AI Search into Firefox for Global Users
Mozilla is rolling out Perplexity as a built-in AI search option in Firefox worldwide on desktop, with mobile support coming soon, after positive tests in the U.S., U.K., and Germany. The integration lets users choose conversational, citation-backed answers from the address bar, and Mozilla says Perplexity won’t share or sell user data while it evaluates adding more AI search options. (AI Insider)
Google Introduces Gemini-Powered “Help Me Schedule” Tool for Gmail and Calendar Users
Google rolled out “Help Me Schedule,” a Gemini-powered tool in Gmail and Google Calendar that reads email context and calendar availability to propose meeting times, drop options into messages, and auto-create invites once confirmed. Built for one-on-one scheduling, it’s part of Google’s broader push to infuse Workspace apps—like Slides, NotebookLM, and Google Vids—with AI-driven productivity features. (AI Insider)
Global Trends in Delivery Robots and Robotaxi Services
The service-robotics market is accelerating, with IFR estimating ~30% growth to ~205,000 units in 2023, APAC taking ~80%, transportation/logistics exceeding half of 2024 deployments, and RaaS contracts up ~31%. Leaders like Starship (8M deliveries), Serve (1,000+ bots), Nuro ($106M pivot to licensing), Waymo (multi-city expansion), and Baidu’s Apollo Go (11M+ rides) are scaling even as Level-4 geofences, safety scrutiny, public trust, and retrenchments (e.g., Amazon Scout) temper momentum heading into 2025–26. (AI Insider)
Research & Innovation
Witnessing Robot Abuse Shapes Human Behavior, Study Finds
Witnessing someone mistreat a service robot can push bystanders in two opposite directions: empathy that leads to helping behavior or behavioral contagion that normalizes incivility, Hanyang University ERICA researchers find. The effect hinges on the robot’s humanlike cues and the observer’s moral identity, suggesting design tweaks and policies—prosocial messaging, ethical conduct guidelines—can encourage respectful human–robot interactions. (AI Insider)
Policy & Governance
California Enacts Nation’s First AI Companion Chatbot Safety Law
California Gov. Gavin Newsom signed SB 243, the nation’s first law regulating AI companion chatbots, requiring platforms like Meta, OpenAI, Character AI, and Replika to implement age checks, self-harm safeguards, clear AI disclosures, limits on sexualized content, and reporting to public health authorities. Taking effect Jan. 1, 2026, the measure follows recent transparency rules and signals California’s push to set the pace on responsible AI governance after incidents involving minors. (AI Insider)
Upcoming Conferences & Events
IEEE/RSJ International Conferenceon Intelligent Robots and Systems
Oct. 19–25 in Hangzhou, China, the 2025 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2025) is a global forum on intelligent robots, covering AI-driven autonomy, perception, manipulation, and human-robot interaction; includes workshops, keynotes, and exhibits. (IROS 2025)
Beneficial AGI (BGI) Summit & Unconference 2025
Oct. 19–23, Istanbul, Turkey and online, the Beneficial AI Summit 2025 returns after last year’s draw of 1,100+ participants from 22 countries. Attendees can expect thought-provoking keynotes, interactive panels, and rich networking focused on AGI, ethics, and innovation. (BGI Summit 2025)
Startups & Capital
JPMorganChase Launches 10-Year, $1.5T Fund Aimed at Strategic Investments Including AI
JPMorgan Chase unveiled a 10-year, $1.5T “Security and Resiliency Initiative,” including up to $10B in direct equity/VC, to boost domestic manufacturing, harden supply chains, and fund frontier tech such as AI, cybersecurity, and quantum. Execution will center on four pillars and leverage new hires, an external advisory council, thematic research, JPM’s scale with ~34k mid-market clients and >90% of the Fortune 500, and policy advocacy on permitting, procurement, R&D, and workforce skills. (AI Insider)
Baselane Announces $34M in New Funding, Debuts AI-Powered Tools to Automate Banking and Bookkeeping for Real Estate Investors
Baselane raised $34.4M—$20M Series B (Thomvest Ventures) plus a $14.4M Series A (Matrix Partners)—and launched “Baselane Smart,” a subscription suite that automates bookkeeping, rent collection, tax reporting, and transfers to save landlords up to 12 hours a week. The company serves 50,000+ investors, has grown 900% since 2022, processes $2B annually, and is positioning itself as a financial hub for U.S. landlords managing 25 million rental units. (AI Insider)
Manchester AI Infrastructure Startup PEAK:AIO Secures £5M to Accelerate Energy-Efficient AI Systems
PEAK:AIO raised £5M to scale its ultra-fast, energy-efficient AI infrastructure, in a round led by Pembroke VCT (£3.7M) with participation from NPIF II — Praetura Equity Finance. The Manchester startup’s next-gen storage, already in use at the NHS, King’s College London, the Zoological Society of London, and Los Alamos National Laboratory, targets AI bottlenecks and will use the capital to expand production to meet global demand. (AI Insider)
Collaborative Fund Leads Phaidra’s $50M+ Series B to Build the AI Factories of the Future
Phaidra raised $50M+ in Series B funding led by Collaborative Fund—with NVIDIA, Sony Innovation Fund, Index Ventures and others—to deploy AI agents that optimize power, cooling, and workloads in large data centers. Founded by ex-DeepMind engineers, the company partners with NVIDIA to maximize compute-per-watt, cutting costs and the environmental footprint of next-gen AI infrastructure. (AI Insider)
Heidi Secures $65M in Series B Funding to Accelerate Building an AI Care Partner for Every Clinician
Heidi raised $65M in Series B funding led by Point72, valuing the healthcare AI startup at $465M and bringing total funding near $100M. Its AI Care Partner—now supporting 73M consults across 116 countries and returning 18M+ clinician hours—is expanding with health systems like the NHS, Beth Israel Lahey, and Monash Health across the U.S., U.K., Canada, Australia, and Europe. (AI Insider)
Radical Ventures Closes $650M Fund to Back Next Generation of AI Startups
Radical Ventures closed a US$650 million early-stage AI fund—among Canada’s largest—adding to a portfolio that includes Cohere and Waabi as global AI startup funding tops US$135 billion annually. The fund includes a US$75 million commitment from CPP Investments (US$280 million total across Radical since 2019), with co-founder Jordan Jacobs saying the firm will keep zeroing in on high-potential founders amid intensifying competition. (AI Insider)
Check out AI Insider every day for the latest in AI and robotics.
