Summary

• Draft EU plans for a Chips Act 2.0 would put more emphasis on demand for European-designed and European-made semiconductors.
• Proposed tools include demand aggregation, offtake agreements, public innovation procurement, and faster approvals for chip facilities.
• The shift recognises that Europe cannot build semiconductor sovereignty through subsidies alone if customers still buy elsewhere.

The European Commission is preparing a renewed semiconductor push that would shift more of Europe’s chip strategy from factory incentives towards demand, procurement, and guaranteed customers for European-designed and European-made components.

A draft plan reported by Reuters and Euronews describes a proposed Chips Act 2.0, expected as part of the EU’s wider technology sovereignty package. The package sits within a broader effort to strengthen Europe’s control over strategic technologies, including semiconductors, cloud, artificial intelligence, and open-source software.

The original European Chips Act focused heavily on manufacturing capacity and supply-chain resilience after pandemic-era shortages exposed Europe’s dependence on chip production in Asia and the US. A demand-led follow-up would acknowledge a tougher commercial problem: new factories do not create sovereignty if European companies and public buyers continue to source critical components elsewhere.

The reported measures include demand accelerators, offtake agreements, a demand forum linking suppliers and users, and greater use of public innovation procurement to support EU-based startups and scaleups. Reuters reported that the Commission document points to €120bn of public and private investment needs by 2035, including about €30bn for advanced semiconductor manufacturing foundry capacity.

The draft also lands against the weaker-than-hoped progress of the first Chips Act. Europe still produces about 10% of global semiconductors, far from the EU’s target of doubling its global market share to 20%. At the same time, AI infrastructure is increasing demand for advanced logic, memory, accelerators, networking components, and power management systems at a pace that stretches supply chains and exposes national dependencies.

A demand-side strategy would mark a more interventionist phase in European industrial policy. Public buyers could be used to support domestic chip startups, strategic sectors could be encouraged to consider European supply, and offtake structures could give manufacturers more confidence to invest. Policymakers frustrated by the slow translation of subsidy programmes into competitive capacity are now looking at the customer side of the market.

The approach carries risk. Semiconductor buyers make decisions on performance, cost, reliability, certification, supply assurance, and integration into existing products. European preference will not compensate for chips that arrive late, cost too much, or fail to meet the requirements of automotive, defence, telecoms, medical technology, energy, and AI infrastructure customers.

Public procurement can build strategic capacity, but it can also shelter weak suppliers when the design is poor. A credible demand strategy would need to distinguish between buying European for resilience and buying European regardless of performance. The first can strengthen the market. The second risks wasting money and slowing adoption.

The geopolitical context is difficult to ignore. The US, China, Taiwan, South Korea, and Japan are all using industrial policy to shape semiconductor capacity. Europe has to act at comparable scale while preserving single market rules, avoiding fragmentation between member states, and supporting both leading-edge chips and the less glamorous components on which industrial systems still depend.

The next test will be the published text. If the Commission turns the draft into a credible mechanism for creating real customers, Europe’s chip strategy will move closer to the commercial reality of the sector. If the package leans on targets without binding demand, the continent may continue to diagnose semiconductor dependency more effectively than it fixes it.