Nvidia allocates $250 million for AI safety research in the United Kingdom

US chipmaker Nvidia announced on Wednesday that it will provide researchers in the United Kingdom with computing resources worth $250 million over the next five years to support projects dedicated to the safety and security of agentic artificial intelligence, according to Reuters.

The initiative involves the annual allocation of computing resources valued at $50 million, aimed at developing safer AI technologies and improving control mechanisms over systems capable of performing complex tasks with a high degree of autonomy.

Some of these resources will be directed to the BOLD laboratory, led by the University of Oxford, SOFAIR, led by University College London (UCL), as well as Imperial College London.

Through this investment, Nvidia aims to support research into the risks associated with advanced AI systems, at a time when the accelerated development of autonomous agents is prompting authorities, technology companies, and academic institutions to pay increasing attention to the security and oversight capabilities of these technologies.

The US company's commitment also reflects the growing importance of computing infrastructure in the development of artificial intelligence, with access to high-performance processors becoming an essential resource for researching and testing next-generation models.

• Nvidia and the London Stock Exchange announce partnership to accelerate capital market data processing

Meanwhile, in the context of the investments announced by Nvidia, London Stock Exchange Group (LSEG), the operator of the London Stock Exchange and one of the leading international providers of financial data and analytics services, announced on Wednesday a partnership with the technology company to accelerate the processing and analysis of large volumes of historical financial market data.

The collaboration aims to enable financial institutions to use computing technologies developed by the US chipmaker to analyze complex information about transactions and financial instrument price movements more quickly.

In recent years, LSEG has expanded its portfolio of services to meet increasingly diverse demand for data and analytical tools, in a context in which artificial intelligence is transforming the way companies access, process, and use financial information. At the same time, the group aims to retain control over its data and the terms of its licensing.

According to the company, analyses have shown that the use of Nvidia's specialized processors allows all trades and price quotes in datasets to be examined in just a few seconds, without the need to create additional copies of the information.

This capability could significantly reduce the time required for financial research and bring historical data analysis closer to real-time processing, at a time when stock trading is moving toward extended operating hours, including 24-hour cycles.

"This collaboration expands the possibilities for clients to extract relevant insights from LSEG's Tick History database, which benefits from robust governance and control mechanisms. Clients want to work quickly with detailed historical data," said Ron Lefferts, co-head of LSEG's Data & Analytics division, according to Reuters.

The British group's Tick History platform contains approximately 110 trillion records covering around 100 million financial instruments, collected over 30 years.

The data comes from 580 exchanges, trading platforms, and financial information providers, making this infrastructure an important resource for analyzing market developments, developing investment strategies, and assessing risks.

The partnership with Nvidia could enable financial institutions to make more efficient use of these volumes of information, reducing the time required for data processing and facilitating the identification of trends and trading opportunities.

The collaboration also highlights the expanding use of high-performance computing infrastructure in the financial services industry, where analytical speed and the ability to process extremely large volumes of data are becoming increasingly important factors in the competitiveness of market participants.