As nations vie for supremacy in critical and emerging technologies, geopolitics and high-tech innovations intersect to shape global dynamics. Technological advances, from artificial intelligence (AI) to space exploration, not only revolutionise industries but redefine global power and strategic alliances. This edition explores how digital payments platforms can be tools of soft power, how Nvidia’s loss in China can be India’s gain, setbacks in NASA’s lunar launches could pave the way for joint missions with India, how Anthropic’s new AI models are affecting cyber risk, and how the Taiwan 2027 invasion could spell the end of cheap tech manufacturing as we know it.
How Digital Payment Platforms are Instruments of Soft Power
The rise of homegrown digital payment ecosystems in major and middle powers has quietly elevated fintech from a commercial phenomenon into a tool of statecraft, giving governments a new and potent channel through which to project influence, shape economic norms, and build networks of dependency.
That is what makes the competition between India’s state-backed payments architecture, China’s commercially dominant platforms, and America’s hybrid ecosystem of private fintech giants and crypto infrastructure so consequential for international relations.
India has constructed its digital payments model on public infrastructure logic. The Unified Payments Interface (UPI), which operates as an open, interoperable system developed under the government’s India Stack initiative, which also encompasses the Aadhaar biometric identification network. The approach is consciously exportable, as the Government of India has made UPI a fixture of almost every bilateral and multilateral engagement, from G20 forums to BRICS summits to the contours of economic cooperation and trade agreements signed with many developed economies and blocs. What India is promoting is not just efficiency; it is a vision of sovereign, interoperable digital cross-border payments that explicitly positions New Delhi against systems that centralise control in foreign commercial or state actors.
China’s approach runs through different channels but aims for comparable strategic results. Alipay and WeChat Pay, the two behemoths of the Chinese payments world, were not created by government decree but have expanded globally with active state backing through the People’s Bank of China. Furthermore, these Chinese platforms are more pragmatic in aiming for retailer level integration in foreign markets instead of government-level or central bank-level agreements only. This makes their penetration a much safer hedge as they can co-exist with similar agreements from other foreign entities and even compete off the back of China’s relatively advantageous cost structure and mass-scaled revenue models.
The US fintech however remains the reigning standard as it has global reach through PayPal, Stripe, and the credit card duopoly of Visa and Mastercard, but Washington has not deployed these the way Beijing or New Delhi deploy their platforms. Binance, once the world’s dominant crypto currency exchange and a de facto piece of American financial soft power projection, has spent years entangled in regulatory disputes. Meanwhile, the prospect of dollar-denominated stablecoins or a digital dollar could recompose this picture, but no coordinated international strategy from Washington has yet emerged.
What Nvidia's Lost Market in China means for India's AI Ambitions
Nvidia CEO Jensen Huang has revealed that his company has virtually zero share in China’s AI chip market in 2026. The loss wasn’t caused by one export restriction but is the outcome of a prolonged policy spiral in which Washington progressively tightened controls, Beijing accelerated domestic alternatives, and Nvidia found itself legislated out of the world’s second-largest economy.
The proximate cause was a series of US government export control measures that restricted the sale of Nvidia’s most capable data centre chips to Chinese buyers on the grounds that they could be used to develop military AI systems. When the cutting edge H100 Graphics Processing Units (GPU) and its successors were blocked from export to China, Nvidia attempted to produce downgraded variants calibrated to fall within permitted thresholds, but those too were eventually restricted or rendered commercially unattractive by further rules.
The H200, more recently permitted under a temporary Trump administration relaxation, has apparently generated no significant revenue because the corresponding Chinese import licences have been regulated by Beijing in a reciprocal manner. Nvidia’s own annual filing with the US SEC acknowledged that its exclusion from China’s data centre computing market had allowed competitors to build larger developer and customer ecosystems to challenge it worldwide, and described the situation as a material threat to its business.
That competitor is principally Huawei, a Chinese tech conglomerate with its finger in every possible pie from consumer electronics to cutting edge computing hardware. The Ascend series of AI processors from Huawei has advanced rapidly under the stimulus of necessity, with Chinese firms that once relied on Nvidia hardware pivoting to domestic alternatives not because they were superior but because they were the only domestic option available. The policy that was intended to deny China AI capability has instead created conditions in which China’s domestic chip industry develops with urgency, scale, and state backing – a pattern where US sanctions and export controls has created competitors and capability through what can only be described as a twisted form of social Darwinism. Huang, likely speaking from a business perspective only, argues that the restriction strategy had largely backfired, noting that ceding an entire market the size of China made limited strategic sense and that the policy needed to be dynamic rather than static.
For India, this reconfiguration of the global AI chip market opens a window that did not exist when Nvidia was simultaneously serving Chinese hyperscalers and American ones. India’s AI Mission, launched under the government’s broader digital economy ambitions, requires compute infrastructure at a scale that has historically been difficult to procure. The programme envisions a network of domestic GPU clusters capable of supporting model training and inference for government agencies, startups, and research institutions. It has set targets for compute capacity that require significant hardware imports, and Nvidia chips are the primary candidate.
With China effectively removed from the market as a Nvidia customer, GPU supply chains face less competition for allocation. Orders that previously would have gone to Chinese data centres now need to find buyers elsewhere. India, which has been negotiating aggressively with Nvidia and other chip suppliers to secure both hardware and preferential pricing as part of AI Mission procurement, is positioned to absorb some of that redirected supply.
However, there are structural caveats. The chips Nvidia is currently permitted to export by the US government and especially in large numbers to most markets are not always the highest-performance variants, and Indian AI research institutions will need access to frontier training hardware to build competitive foundation models rather than merely deploying existing ones. The gap between inference-grade compute and training-grade compute is large, and closing it requires either access to Nvidia’s most capable chips or the development of viable domestic alternatives. India’s homegrown chip design efforts remain nascent compared to those in the US or Taiwan.
Blue Origin and SpaceX Mishaps and the Path forward for NASA, ISRO, and the Lunar Crewed Missions
A static test fire of Blue Origin’s New Glenn rocket in May 2026 ended in a fireball. While there were no human casualties, the explosion may significantly alter the timeline of the US’s Artemis programme, and the credibility of a commercial partner that NASA has staked a significant portion of its lunar ambitions upon.
At the time of the accident, New Glenn’s fourth mission, the NG-4 flight, had only been cleared to proceed a week earlier after the US Federal Aviation Administration (FAA) wrapped up its investigation into an earlier anomaly during the NG-3 launch in April, when a second-stage failure had stranded a satellite in an unstable orbit. That investigation had already cost weeks of schedule; now the erector launcher is gone and the launch pad is going to take at least a few months or more to get back into shape.
NASA selected Blue Origin alongside SpaceX to provide the lunar landers that will carry astronauts to the moon’s surface as part of Artemis 3 and subsequent missions. Artemis 3, already pushed to late 2027, was conceived as a crewed mission that would rendezvous with a lunar lander in orbit and deliver astronauts to the surface. The programme has now essentially been reduced to a single working lander option in the near term. SpaceX’s Starship too experienced a first-stage failure during its twelfth test flight in May 2026, which further complicates the reliability picture for the entire Artemis programme.
For India, this turbulence in the American lunar programme creates an opening for a deepened partnership with NASA that goes beyond the existing Artemis Accords framework. ISRO’s own crewed spaceflight programme, the Gaganyaan programme, has been delayed multiple times, and India has not yet achieved the human lunar ambition it has periodically articulated. However, ISRO’s track record in cost-efficient lunar surface operations, demonstrated by Chandrayaan-3’s South Pole landing in 2023 and its growing capability in heavy lift through the LVM3 rocket, makes it a plausible partner for science payloads, robotic precursor missions, and eventually joint crewed operations.
Claude Mythos and Fabel in the New Era of Cyber Risk
US AI giant Anthropic has been forced to curtail the user base of its most advanced model Fable, and its more capable version Mythos, after a US Government directive. However, there are reports that at least a version of Fable 5 had been jail-broken while it was available in early June 2026, which means the guardrails on its capabilities programmed to disable harmful use of the model in unethical or illicit purposes, were circumvented.
However this was far from the first controversy the model was involved in. Earlier in 2026, Anthropic had confirmed that the model had identified thousands of severe security flaws across major operating systems and browsers, including vulnerabilities that had been dormant for decades without human detection. The implications of that finding had a near apocalyptic projection for the digitally enabled global financial system and the operational technology controlling critical infrastructure. The vulnerabilities in these systems that Mythos reportedly found were essentially what penetration testers and cyber security professionals call ‘0 Day Vulnerabilities’, which are bought and sold as cyber weapons by State and non-State actors illicitly as ways to compromise targeted digital systems. 0 Day Vulnerabilities by their very nature are patched only after they are exploited by attackers as no amount of code hygiene and manual audits can be thorough enough to catch everything. However, Mythos demonstrated that a sufficiently capable AI model can find them in hours.
What Mythos and the new paradigm represents for the longer-term relationship between AI capability and cyber security is a fundamental inversion of the previous equilibrium. The gap between what an AI-equipped attacker can do and what a human-staffed security operations centre can detect in real time is now large enough to constitute a structural risk.
How a Taiwan 2027 Invasion Scenario could End the Age of Cheap Tech
The concentration of global technology manufacturing in Taiwan and China was not an accident of geography; it was the cumulative result of deliberate corporate and policy choices made over decades, optimised for cost and shareholder returns with little regard for strategic exposure. The consequences of those choices are now structural, and the window to reverse them before a potential Taiwan conflict in 2027 is effectively closed.
Taiwan’s chip conglomerate TSMC produces a majority of the world’s advanced logic chips and its customers include leading US companies such as Apple, Nvidia, AMD, and Qualcomm. No other company, not South Korea’s Samsung, not US-Israeli Intel, not any emerging foundry, can manufacture at comparable volume and node sophistication. TSMC’s position was built through sustained state investment in technical education, deliberate industrial clustering, and two decades of accumulated manufacturing knowledge that competitors cannot replicate on a compressed timeline. The broader Taiwanese ecosystem compounds this dependence, including substrate manufacturers, advanced packaging firms, and specialised materials suppliers whose outputs are necessary preconditions for TSMC’s own production.
China occupies the tier below; its factories produce a dominant share of passive components, printed circuit boards, connectors, power management chips, display panels, and electromechanical subassemblies that go into virtually every electronic product worldwide. This is not simply cheap labour; Chinese manufacturers have built process expertise, supplier networks, tooling knowledge, and logistics infrastructure over decades that Vietnam, India, and Mexico cannot absorb on any near-term timeline, regardless of capital availability.
The US and Japan deliberately let this happen because the firms doing it were profitable, as American semiconductor companies went fabless, outsourcing production to TSMC while retaining design margins. Japanese electronics manufacturers retreated from volume production into premium industrial and automotive niches, ceding the components market to lower-cost competitors. In both cases the corporate logic was sound but the strategic logic was lacking.
In the event of TSMC’s production capabilities going offline or any war related disruptions in China’s Shenzhen and Chengdu high-tech industrial base would effectively mean the era of cost-optimised globalised tech manufacturing will end. What follows will likely be more expensive and fragmented, and likely regress global standards of living to unthinkable levels.
Check these out
- Pope Leo, Magnifica humanitas’: AI must serve humanity not concentrate power. The Vatican. URL: https://www.vatican.va/content/leo-xiv/en/encyclicals/documents/20260515-magnifica-humanitas.html
- Lily Hay Newman, ‘Amazon Is Using Specialized AI Agents for Deep Bug Hunting’. Wired, 24 November 2025. URL: https://www.wired.com/story/amazon-autonomous-threat-analysis
- Jon Peddie, ‘AMD's new MI325X Instinct AI GPU’. Jon Peddie Research, 10 October 2024. URL: https://www.jonpeddie.com/news/amds-new-mi325x-instinct-ai-gpu
- Sarah Sewall and Ming Lou, ‘The Geopolitics of Digital Currency’. Harvard Kennedy School Belfer Center, 21 January 2022. URL: https://www.belfercenter.org/publication/geopolitics-digital-currency
- Leonard David, ‘China shakes up its space programs to land astronauts on the moon by 2030: “We will spare no effort”'. Space Insider, 27 May 2026. URL: https://www.space.com/space-exploration/china-shakes-up-its-space-programs-to-land-astronauts-on-the-moon-by-2030-we-will-spare-no-effort
- ‘Buddhism and AI: Another Look’. The Digital Orientalist. URL: https://digitalorientalist.com/2026/03/24/buddhism-and-ai-another-look/
