Investigating the Potential of Pax Silica
A Technology Grouping for the AI Age
Authors
Sulaiman Akhtar of SPAN, a publication of the U.S. Embassy in New Delhi, spoke to me about the geopolitical implications of Pax Silica. I am reproducing my answers to his questions below.
You can read the interview in SPAN here
Q1: To begin with, what is ‘Pax Silica,’ and why did the United States launch this initiative for the global semiconductor ecosystem? I would first correct a common misconception: Pax Silica is not just about semiconductors. Reading the declaration makes it clear that this is a technology grouping for the AI age. The name references silicon, but the scope spans the entire technology stack: energy, critical minerals, advanced manufacturing, semiconductors, compute, frontier foundation models, software platforms, information connectivity, fibre-optic cables, data centres, and transportation logistics. The declaration explicitly says it covers “strategic stacks of the global technology supply chain, including, but not limited to”, these segments. Semiconductors are one critical layer, but Pax Silica is really about who controls the infrastructure of the AI economy.
The “Pax” framing is deliberate. It is the US signalling that it intends to establish a durable economic and technological order for the AI era, anchored in a coalition of like-minded partners.
Now, semiconductors remain central to this broader initiative. In our book “When the Chips Are Down,” my co-author Abhiram Manchi and I propose that semiconductors are not merely a “critical” technology; they are “metacritical.” This metacriticality arises from the convergence of three distinct dimensions.
The first is geopolitical. Semiconductors have become a key domain in the US-China confrontation. The US has imposed export controls restricting Chinese access to essential equipment, software, and intellectual property, while the Taiwan Strait situation has made the concentration of advanced manufacturing on that island a global security concern.
The second dimension is geoeconomic. COVID-19 exposed how the hyper-globalised semiconductor supply chain had traded resilience for specialisation and cost-effectiveness, and the dominance of single players in highly specialised steps means that a delay from one company can halt production at thousands of downstream firms.
The third dimension is technological. Semiconductors are foundational for virtually any technological advancement. If you want better military hardware, get specialised chips; if you want to dominate the AI race, get better AI chips; if you want next-generation communications, get new baseband chips. The lack of a dependable chip supply has downstream effects on every future technology.
The US launched Pax Silica because it knows that no country can be self-sufficient in this domain. Multilateral cooperation between capable partners is a necessity, not a choice. This realisation has driven Washington to seek an allied framework that goes well beyond chips to encompass the full minerals-to-models pipeline. The US Under Secretary of State Jacob Helberg has also said that Pax Silica stands for the rejection of ‘weaponised dependency.’
That said, the current fact sheet is heavy on vague announcements but light on executable policy. We should assess the initiative by what it delivers operationally, not by the ambition of its rhetoric.
Q2: What are the key vulnerabilities or risks in today’s global semiconductor supply chain that Pax Silica seeks to address? In our book, we identify three geoeconomic risks that make the semiconductor supply chain susceptible to disruption. These risks sit at the heart of what Pax Silica seeks to address.
First, geographic concentration. East Asia has a heavy concentration of manufacturing and testing bases. Taiwan alone accounted for about 80 per cent of global contract foundry revenues. The result is that even a localised health, political, or social crisis can halt all industries utilising semiconductor chips.
Second, single-point-of-failure dominance. In several highly specialised steps across the supply chain, just one global player dominates. ASML in the Netherlands is the sole manufacturer of extreme ultraviolet lithography machines. The three EDA tool makers — Cadence, Synopsys, and Mentor Graphics — hold approximately 97 per cent market share, and all are American. A delay from one such company can hit the production roadmaps of thousands of downstream firms. These are not just commercial risks, but chokepoints that can be weaponised.
Third, and this is what makes the supply chain a geopolitical flashpoint: several of these bottlenecks make semiconductors an attractive tool of statecraft. Nation-states can try to block one or more of these chokepoints for political purposes. This is precisely what the US has done with its export controls on China, and what China has done with its export restrictions on critical minerals.
Q3: How does Pax Silica address these risks differently from earlier supply-chain strategies that focused mainly on efficiency and cost? For decades, the semiconductor industry was the poster child of globalisation. As we describe in our book, this complex supply chain ran like a well-oiled machine: the US and Europe led on chip architectural design; Indian design centres implemented these designs; Taiwan and South Korea took care of manufacturing; China, Singapore, and Malaysia were involved in packaging; and Japanese firms excelled in specialised manufacturing materials. The system was optimised relentlessly for efficiency and cost. Nobody seriously worried about the geopolitical implications of this hyper-specialisation.
Pax Silica represents a fundamental shift. It is not an efficiency play. It is a security-and-prosperity play. The core idea is that some economic efficiency must be sacrificed to build resilience, reduce single-source dependencies, and ensure that the infrastructure of the AI economy is controlled by trusted partners.
Earlier efforts were largely national in scope. The US CHIPS and Science Act, Japan’s incentive programmes, the EU Chips Act — each country is trying to build domestic capacity on its own terms. In our book, we argue that these purely national industrial policies have several structural flaws. The current rush for advanced fabs is driven by government concerns rather than market demand. Chips produced outside countries with a comparative advantage are costlier. For example, TSMC has repeatedly highlighted that building fabs in the US is far more expensive than in Taiwan. And the current level of subsidies will not be enough; companies will demand evergreening of subsidies to keep higher-cost, underutilised fabs running. At some point, this becomes unsustainable.
Pax Silica attempts to address these flaws by moving from unilateral industrial policy to coordinated, multinational action, and covering not just chips but the entire stack from minerals and energy through compute to AI models and logistics. The idea is not that each country replicates the whole chain, but that trusted partners specialise in complementary segments across the full stack, with enough redundancy to withstand disruption. The hope is that the grouping will collectively, if not individually, have capabilities across all supply chain stages.
In the old paradigm, the cheapest chip won. In the Pax Silica framework, the trusted technology wins. Trusted in the sense that its provenance, from raw mineral to deployed AI system, passes through aligned nations. The declaration talks about “trusted information networks” and a “shared and trusted ecosystem of AI developers and vendors.”
However, the tension between resilience and fragmentation is real. If Pax Silica becomes a tool for creating parallel, bloc-based technology ecosystems rather than diversified global ones, it could increase costs and reduce innovation without meaningfully improving security. Our book’s concluding argument is that globalisation and cooperation will ultimately outlast aggressive national competition. Countries will have to come to a new modus vivendi on semiconductors. A complete decoupling seems highly unlikely. Pax Silica will have to navigate this tension carefully.
Q4: In simple terms, how does Pax Silica aim to make semiconductor supply chains more secure, diversified, and trustworthy without fragmenting the global market?
This is the central tension, and I am not sure the initiative has fully resolved it yet.
In principle, Pax Silica seeks to create a “trusted zone” across the full AI-age technology stack. Within this zone, aligned nations would allow a relatively free flow of minerals, energy cooperation, manufacturing capacity, chip design IP, AI models, digital infrastructure, and data. The members collectively cover enough of the stack that no external actor can exercise coercive leverage over any single segment.
The diversification piece works by distributing capability across multiple allied geographies at each layer. If one mineral source is disrupted, alternative partner sources can step in. If one manufacturing node faces a contingency, others can absorb some of the load.
The trustworthiness dimension is about supply chain provenance across the entire stack. The declaration talks about protecting “sensitive technologies and critical infrastructure from undue access, influence, or control.”
This approach maps well onto one of the four strategic objectives we identify in our “siliconcraft” framework: increasing supply chain resilience. It acknowledges that the semiconductor industry’s well-being is crucial to the Information Age and that cooperation is required. The tools for achieving this include promoting open-source hardware, mapping supply chains to identify vulnerabilities and forecast shocks, and building partnerships with countries that have complementary capabilities.
Now, the fragmentation risk. The semiconductor supply chain is extraordinarily capital-intensive and knowledge-intensive. In our book, we note that technology production networks are global now, unlike the disjointed ecosystems during the Cold War. Developing isolated science and technology ecosystems is no longer possible. Duplicating the leading edge across blocs may be physically impossible within any reasonable timeframe. The market for EUV machines, advanced photoresists, and the most sophisticated AI training clusters involves natural monopolies or near-monopolies.
Our concluding argument in the book is relevant here. This round of semiconductor geopolitics will possibly end with a slightly more diversified supply chain and a realisation that semiconductor interdependence is a boon, not a bane. Pax Silica can meaningfully improve resilience across the mid-range and trailing-edge segments and create useful redundancy in areas like mineral processing, energy, and digital infrastructure. At the leading edge, the initiative will struggle to avoid some degree of fragmentation — or will remain aspirational.
Q5: Where does India fit into the Pax Silica framework, particularly in areas such as chip design, manufacturing, and talent development? India’s relationship with Pax Silica needs to be understood at two levels: diplomatic and structural. At the diplomatic level, India was absent from the initial announcement, which triggered alarm. But I argued from the outset that this matters far less than the noise suggested. By January 2026, the US envoy invited India to join. We have seen this movie before. India was not a founding member of the Mineral Security Partnership either and joined a year later.
At the structural level, India is already deeply embedded across multiple layers of the technology stack that Pax Silica seeks to secure.
Start with chip design. This is where India’s strength is most formidable. As we document in our book, semiconductor design is where India’s comparative advantage lies. With an estimated 20 per cent of the world’s design engineers and design centres of the top American semiconductor firms located in the country, India possesses a critical mass of talent. Nearly 3,000 chips are being designed annually by 30,000 engineers. Although India started as a software and support hub, its design houses have climbed the technology ladder, embedding themselves firmly in the global value chain. The chips powering Pax Silica are being designed in Bengaluru, Hyderabad, and Noida.
However, our book also identifies a critical weakness: not much IP is owned by Indian companies. The cumulative revenue of domestic semiconductor design companies is relatively small. This is problematic because countries registering the IP can restrict Indian players in a geopolitical kerfuffle. India has the talent, but it needs to translate that into Indian-owned, globally competitive fabless companies with their own IP.
On manufacturing, eight Outsourced Assembly and Test (OSAT) plants, one compound semiconductor fab and one commercial silicon fab are under construction. On talent, this is where I believe India’s quickest route to strategic leverage lies. Our book argues that if Indian talent becomes a dominant driver of the global semiconductor supply chain, interdependence on chips with other countries will not be a cause for concern. The US itself faces a shortfall of 300,000 engineers and 90,000 skilled technicians by 2030. India’s talent is in enormous demand.
Beyond semiconductors, India is relevant to every layer of the Pax Silica stack. Indian software firms are positioned to become the global delivery vehicle for enterprise AI. India is one of the largest markets for AI deployment. It has critical mineral reserves and growing processing capacity. Its digital infrastructure market is scaling rapidly.
Q6: How can closer U.S.–India cooperation in semiconductors strengthen economic resilience and reduce shared strategic risks? The US-India technology relationship is already deeper than most people realise. The hundreds of Global Capability Centres that American chip companies operate in India are integral to their innovation pipelines. This creates a natural foundation for cooperation that maps onto the broader Pax Silica framework.
Since no country can be self-sufficient, governments must find trusted partners with complementary strengths. The US-India relationship is a natural fit because the complementarities are deep: the US leads in chip design tools, EDA software, equipment, and frontier AI models; India provides design talent at scale, enterprise AI integration capability, and an enormous market.
There are several concrete pathways. On design and IP, closer cooperation could involve joint R&D on next-generation chip architectures, including AI-specific chips, RISC-V-based designs, and domain-specific accelerators. Our book discusses how the open-source RISC-V architecture and initiatives like OpenROAD could reduce entry barriers and EDA costs. India’s large design workforce could benefit enormously from such cooperation, and the US benefits from having more trusted partners contributing to the innovation pipeline.
On supply chain diversification, the US wants to reduce dependence on any single manufacturing geography. India offers a viable alternative for trailing-edge and mid-range chip production and assembly. The book argues that over a two-decade period, sequenced investments, starting with OSAT, then trailing-edge fabs, then gradually more advanced facilities, can build India into a credible manufacturing node.
On the AI deployment layer, Indian software firms are positioned for what Kai-fu Lee calls “enterprise AI.” The Pax Silica declaration talks about a “trusted ecosystem of AI developers and vendors to renew legacy industries and unlock new markets.” India is where much of that renewal will actually happen. Any US strategy to shape the global AI economy that bypasses India’s integrators and market is incomplete.
On critical minerals and energy, India has reserves and is building processing and recycling capacity. India’s enormous and growing energy demand also makes it a natural partner for the energy cooperation dimension of Pax Silica.
The shared strategic risk is also clear. Neither India nor the US would want China to gain disproportionate influence over technology supply chains.
Q7: What role do policy coordination and trusted partnerships play in making Pax Silica effective? Policy coordination is the make-or-break factor. Building resilience requires mapping supply chains across countries to identify vulnerabilities and forecast shocks, harmonising standards so that “trusted” chips and AI systems from one member are accepted in others, aligning export controls to prevent arbitrage, coordinating investment screening against non-allied acquisitions, and promoting open-source hardware to reduce barriers to entry.
The “trusted partnerships” framing is important because it distinguishes Pax Silica from a purely transactional arrangement. As long as the semiconductor supply chain primarily resides among trusted partners, interdependence is not a strategic vulnerability.
However, I have reservations about execution. First, the initial coalition has notable gaps. Taiwan, the most critical manufacturing node, is absent. India, the largest design talent pool and a massive market, was initially excluded. Major European economies are not included yet. Without these pillars, the initiative risks being a paper tiger.
Second, policy coordination among democracies is inherently difficult. Each country has domestic interests, lobbying ecosystems, and electoral pressures. Governments face pressure to show “national” results from international cooperation.
Q8: Looking ahead, how could sustained U.S.–India alignment in semiconductors shape technological leadership and strategic trust over the next decade? Let me offer both an optimistic scenario and the conditions that must hold. And let me frame this in terms of the full Pax Silica stack, because that is where the real opportunity lies.
In our book, we argue that the current round of semiconductor geopolitics will likely end not with complete decoupling but with a slightly more diversified supply chain and a realisation that interdependence among trusted partners is a boon, not a bane. If that prediction holds, the US-India relationship could become one of the central pillars of this new equilibrium.
In the optimistic scenario, India’s chip design talent grows and produces more Indian-owned fabless companies with globally competitive products and their own IP. India’s OSAT and fab infrastructure reaches operational maturity, giving the alliance a credible alternative manufacturing node. Indian firms become the leading integrators of enterprise AI globally. India’s data centre and fibre-optic infrastructure scales into a trusted Pax Silica node. And India’s critical mineral processing capacity matures.
This would mean the democratic world’s technology stack runs substantially through US-India channels, reducing the leverage any single actor can exercise. The trust built through deep technology cooperation would spill over into broader strategic trust across defence, space, and governance. Our book discusses how the US has historically been willing to share sensitive technologies with partners when geopolitical competition demands it, AUKUS being a recent example. Sustained semiconductor cooperation could place India in a similar category.
Views are personal and don’t necessarily represent those of the institution