How the UK can compete in AI, changing seasons & more
The Oxford China Policy Lab Newsletter
Hi folks,
For centuries, September has been a month that has stirred something deep in the hearts of British poets, authors and artists. For John Keats, it marked the “season of mists and mellow fruitfulness”; William Shakespeare used autumn as an allegory to explore his own ageing: “That time of year thou mayst in me behold / When yellow leaves, or none, or few, do hang / Upon those boughs”; Emily Brontë added “Every leaf speaks bliss to me / Fluttering from the autumn tree.”
I find it has a wistful energy: look up, and you can see the end of the year peeking its head over the horizon, beckoning us closer with shortening days, browning leaves, and colder nights. But September is also the start of the academic year, and the return of Parliament, following a long hot summer of local politics.
Like old leaves blowing off a shaking tree, September saw one of the most consequential reshuffles yet in Keir Starmer’s premiership. Spurred by the need to replace his deputy, the Prime Minister ended up moving dozens of senior and junior officials into new positions in or out of government. The senior ones will now be expected to balance their Downing Street orders with their own visions and ambitions for what their role can achieve: how this ends up translating in reality remains to be seen. Their junior counterparts will be keen to impress, and will be working out how to continue their rise up the system. From an analyst point of view, this now means trying to ascertain what the consistencies will be, and to what degree each of these new ministers can exert their own influences in the role. On a macro level, will this be the same set of people in power in a year’s time? Will they make it through the cold political months ahead? Who will bloom first when the next change of seasons arrives? These questions matter for both geopolitical and emerging technology assessments.
Ultimately it’s too early to tell if the reshuffle will have a meaningful impact on the Government’s China or artificial intelligence approaches. In the case of the former, the new Business and Trade Secretary Peter Kyle was in Beijing within a week of being reshuffled (via Washington D.C. first). In the latter, President Donald Trump’s second state visit days after the reshuffle saw substantial investment and deals signed between US and UK technology companies, with US firms from Google to Palantir pledging billions for the UK tech sector. The visit also resulted in further agreements for both nations’ AI institutes to collaborate. But both of these plans were in the works long before the reshuffle, so they shouldn’t be read as signals beyond that.
With that being said, one area of consistent strength for the UK’s AI and diplomatic positioning is its AI Security Institute, which we highlighted in a letter published by The Economist this week.
Seasons change, but there is always continuity between them. This exists within OCPL too: my colleagues have met and briefed dozens of diplomats, officials and leaders on complex thematic issues in a clear and succinct manner across a number of different countries over the last year. And as September draws to an end, we’re showing no signs of slowing down. Autumn will bring new ideas in new shapes, new fellowship applications, and an original newsletter project focusing on our research. Watch this space.
Sam Hogg
Head of Policy Engagement
News and Views
The research and analysis of OCPL experts are their own and do not reflect organisational views
Oxford China Policy Lab expert analysis has again been published by the British Parliament. In evidence for Parliament’s Science, Innovation and Technology Committee, Kayla Blomquist and Sam Hogg argue that the UK’s current posture and strategic approach to artificial intelligence give it significant soft power influence and leverage in the scientific realm.
Kayla Blomquist and Sam Hogg were published in The Economist. In the leading letter, they argue that Britain can export expertise in AI auditing, quality evaluation and AI’s responsible deployment while reinforcing its position as an essential mediator between AI superpowers.
Kayla Blomquist moderated a panel for Young China Watchers Hong Kong on AI in the Corridors of Power: Can LLMs Reshape Policy and Diplomacy? with Joseph Castellano (Anthropic), Alex Duffy (Every Inc.), and Jinghan Zeng (City University of Hong Kong). The discussion examined how LLMs are beginning to influence diplomatic practice, while highlighting the enduring importance of human judgment, cultural nuance, and transparency.
Scott Singer published a commentary for the Carnegie Endowment on China’s new “AI+” strategy, arguing that structural economic and investment constraints make Beijing’s goal of diffusing AI across 90% of its economy by 2030 unlikely to succeed.
What We’re Reading
“An Open Door: AI Innovation in the Global South amid Geostrategic Competition” by Noam Unger and Madeleine McLean. This CSIS report examines how AI adoption in low- and middle-income countries is shaped by access to compute, cost barriers, and governance capacity, warning that constraints risk pushing them toward Chinese alternatives and proposing U.S. steps to promote open and democratic AI pathways.
“Full Stack: China’s Evolving Industrial Policy for AI” by Kyle Chan, Gregory Smith, Jimmy Goodrich, Gerard DiPippo, and Konstantin F. Pilz. This RAND study analyzes Beijing’s full-stack approach to AI industrial policy, spanning chips, compute, software, and applications, highlighting how state funds and labs are accelerating adoption while U.S. export controls, weak private investment, and immature domestic hardware still constrain progress.
“China’s Manufacturing Innovation Centers: A Benchmarking Report with the Manufacturing USA Network” by Michael Molnar et al. This NIST report benchmarks China’s 33 state-led manufacturing innovation centers against U.S. and German models, underscoring how Beijing’s scale of public financing, equity-style institutes, and state-directed coordination make the MICs central to its advanced manufacturing and techno-industrial strategy.
“A Third Path For AI Beyond The US-China Binary” by Dang Nguyen. This essay argues that a “third stack” of AI is emerging beyond the U.S.–China binary, where countries pursue infrastructural nonalignment by assembling open-weight models, sovereign clouds, local data regimes and bespoke licensing to exercise “epistemic sovereignty.”