Investors are on edge as talk of an AI bubble gets louder (www.thetimes.com)

A wave of froth around AI investing is making investors nervous: a lavish Balderton celebration underscored a wider narrative highlighted by PitchBook data showing extreme concentration of private-market capital. More than 40% of US venture funding this year flowed into just ten companies, while OpenAI, Anthropic and Databricks alone have absorbed north of $100 billion. Headline rounds — Databricks’s $1B at a $100B valuation (up 61% from Dec 2024) and Anthropic’s $13B Series F valuing it at $183B (roughly triple six months earlier) — sit alongside record-high AI funding in the first nine months of 2025 and unusually rapid deal timelines. The technical and market implications matter for the AI/ML community. Money is piling into late-stage, capital-intensive models and infrastructure, raising the bar for performance and scale required to compete, while early-stage funding and VC fundraising are drying up (US fundraising near 2017 lows; UK VC inflows about £7B, down ~40% from 2024). Limited partners are hesitant to reinvest amid weak exits and muted IPO activity, trapping capital and creating thin liquidity. That concentration boosts short-term capability for big-model progress but heightens systemic risk: if sentiment flips, startups could face harsher funding conditions, slower hiring, and an elevated expectation gap between valuation and deliverable AI product-market fit.