London-based AI start-up Basecamp Research has raised $140mn from investors including Anthropic, Nvidia and the UK’s sovereign AI fund as it seeks to build a new generation of artificial intelligence models trained on DNA rather than language. The company’s Eden models run on its proprietary “Trillion Gene Atlas”, which it describes as the world’s largest biological AI training dataset, built from genetic data collected through around 150 expeditions across the world.

Basecamp argues that a shortage of high-quality biological data is becoming a constraint on AI-driven drug discovery. Co-founder and chief executive Glen Gowers described the problem as the absence of “internet-scale data” in biology, with the company collecting genetic information from environments ranging from the Arctic and Chile to oceans and volcanoes. The latest funding will expand research into regions including Oceania and the high Arctic, while supporting partnerships in countries such as Cameroon, Malawi and Chile. Basecamp is initially focusing on cell therapies, with two AI-designed treatments already showing promising results in mice, according to the company.

The funding highlights a broader shift in AI investment towards specialised datasets rather than computing power alone. Basecamp said its valuation is now “very close” to $1bn, while the sovereign AI fund’s participation makes it the sixth company in which the £500mn state-backed vehicle has invested directly. Lord Vallance, chairman of the government’s AI taskforce, said AI could accelerate scientific discovery and make advanced therapies simpler to deliver. As AI moves deeper into healthcare, access to proprietary biological data could become as strategically important as access to chips and computing infrastructure.

- Muthu Ramanathan

Growing concerns over the risks of artificial intelligence are beginning to raise questions about the sustainability of the investment boom supporting the technology. Anthropic chief executive Dario Amodei has called on leading AI companies to “slow the pace” of development, gaining support from OpenAI chief executive Sam Altman and Elon Musk, although Nvidia chief Jensen Huang and Meta founder Mark Zuckerberg have been less receptive. The debate has already translated into sharp market movements, with almost $700bn wiped from a basket of 25 AI-related stocks last Monday before $169bn was recovered by the end of the week.

The potential economic impact extends well beyond AI model developers. Panmure Liberum head of research Joachim Klement warned that a slowdown could trigger “a big correction in stock markets worldwide”, partly because frontier AI development is driving enormous infrastructure spending. The five major hyperscalers are forecast to spend $5tn on AI infrastructure over the next five years, supporting semiconductor manufacturers, electrical equipment companies, data-centre operators and property owners. Anthropic and OpenAI could also face pressure because of take-or-pay computing contracts that require payment for capacity even if it is not ultimately used.

However, a slowdown in frontier models does not necessarily mean the end of the wider AI investment cycle. Anthropic’s annualised recurring revenue reached $65bn in July, up from $9bn at the start of the year, while Deutsche Bank’s Adrian Cox argues that enterprise adoption remains a more important driver of infrastructure demand than advances at the frontier. Around 17% of UK businesses currently use AI, leaving substantial room for adoption to expand. The key question for markets is therefore whether AI spending can increasingly be supported by commercial demand rather than expectations surrounding ever more powerful models.

- Muthu Ramanathan

Unions, environmental groups and think tanks are calling for the UK’s National Wealth Fund to be given greater financial firepower, arguing that a larger public investment institution could help accelerate infrastructure, industrial and regional investment. The coalition wants the NWF to become a “world-leading national development bank”, with greater independence to raise finance and invest over longer time horizons. The fund was launched in July 2024 to attract private capital into projects including ports, gigafactories, hydrogen and steel.

The scale of the proposed expansion reflects the gap between the UK’s public investment capacity and that of comparable economies. The NWF currently has around £5.5bn available to invest each year for the next five years, compared with €62bn lent by Germany’s state-backed KfW in 2025 alone. Supporters argue that greater financial capacity could fund home retrofitting, take stakes in critical infrastructure, support green industries in areas affected by de-industrialisation and establish regional banks for small businesses. They also argue that greater borrowing capacity could remain consistent with the government’s fiscal rules if funding is directed towards investment rather than day-to-day spending.

The government points to the NWF’s existing record, saying it delivered £3.9bn of investment during its first year and generated a further £5.25bn of private finance, while supporting 11,500 jobs. The Treasury says the fund remains focused on “crowding in private capital” and driving growth across the UK. The debate therefore centres on whether the NWF should remain primarily a catalyst for private investment or develop into a much larger source of public finance. Its future scale could determine how effectively the UK converts limited public resources into long-term infrastructure and productivity gains.

- Shrish Yalamarti

Until next time,

The Long and Short

Markets, Made Sense