BenevolentAI co-founder Ken Mulvany has called on the UK government to scrap stamp duty on share trading to help British technology companies secure investment.
The 58-year-old artificial intelligence pioneer said the current tax regime discourages backing for home-grown businesses and risks driving top entrepreneurs to the United States.
Writing for publication ThisIsMoney, the sister outlet of The Mail on Sunday, Mulvany urged ministers to reform investment taxes, overhaul the planning system, and reduce industrial electricity costs to support economic growth.
Britain currently imposes a 0.5 per cent stamp duty tax on purchases of UK shares, a cost that does not apply to most overseas equity purchases. Mulvany argued that reducing this financial disadvantage would encourage major domestic investment funds to keep capital within the UK.
Tax reform and venture capital
"Britain imposes a 0.5 per cent tax on buying UK shares while most overseas shares can be bought without it," Mulvany wrote. "Reducing that disadvantage would send a powerful signal about where we want capital to go."
Mulvany warned that without changes to investment tax rules, promising British technology startups will continue migrating to established American venture capital hubs such as Boston or San Francisco during early funding rounds.
"The next generation of AI companies may need few people, but they will still need money to grow," Mulvany said, "and Britain cannot go on producing excellent science and ambitious founders only to watch their businesses leave for Boston or San Francisco at the first serious funding round."
"If we want British savers to own a meaningful share of companies built on cheap intelligence, the tax treatment of investment must make it worth their while," he added.
Stamp duty on shares in the United Kingdom is a tax charged on paperless share transactions of UK-registered companies. In contrast, international equity markets such as the New York Stock Exchange and Nasdaq do not levy equivalent transaction duties on equity purchases, giving foreign stock exchanges a competitive cost edge when attracting capital.
Energy costs and data centre planning
Beyond share taxes, Mulvany warned that soaring industrial electricity prices and sluggish planning approvals threaten the expansion of artificial intelligence infrastructure across the country.
While acknowledging that Britain lacks the hundreds of billions of pounds being invested in artificial intelligence by the United States or China, he stressed that the UK possesses scientific expertise, established financial markets, and world-class universities that put it in a prime position to benefit as computing technology becomes cheaper.
However, high energy costs remain a major obstacle. "Britain has some of the most expensive industrial electricity in the developed world and a grid never designed for the demand now arriving," Mulvany said.
Mulvany also highlighted bureaucratic delays in obtaining planning permission for data centres required to host high-performance computing hardware. "The machinery in a data centre is obsolete in three years," he said. "A planning process that takes three years is therefore not a delay but a veto. What is needed is a presumption in favour."
Data centres rely on massive amounts of electrical power to run servers and cooling equipment for complex artificial intelligence calculations. Modern machine learning facilities require substantial grid connections, but complex local planning procedures in the UK often take several years to secure development approvals.
Drug discovery and BenevolentAI background
Mulvany was an early proponent of artificial intelligence as a tool for economic productivity, advocating its potential long before the public emergence of major commercial laboratories such as OpenAI and Anthropic.
He co-founded BenevolentAI in 2013 with the goal of deploying artificial intelligence algorithms to process vast quantities of clinical data and accelerate pharmaceutical development.
Headquartered in London, BenevolentAI has partnered with global pharmaceutical companies, including AstraZeneca. In 2020, the company used its technology platform to identify Baricitinib, an existing medication for rheumatoid arthritis, as a candidate treatment for Covid-19.
OpenAI and Anthropic are leading American artificial intelligence developers known for large-scale AI models. AstraZeneca is a British-Swedish pharmaceutical giant headquartered in Cambridge. Baricitinib, originally developed for joint inflammation, subsequently received emergency treatment authorisations from global health regulators during the coronavirus pandemic following clinical evaluation.
Mulvany concluded that ministers must tackle tax, energy, and planning barriers promptly so that domestic investors can benefit from the productivity surge promised by emerging AI technologies.
