Political forces are now intersecting with the push for advanced artificial intelligence. After the recent US midterm elections, UBS CEO Sergio Ermotti cautioned that uncertainty in politics, along with possible stricter rules, might slow the huge infrastructure buildout needed for next-generation LLMs. His remarks mark a shift away from unchecked excitement toward a clearer look at data center construction, power supplies, and hyperscaler spending.
Quick Take
- What happened: UBS Chief Executive Sergio Ermotti stated that the US midterm election results—specifically a Democratic sweep—could introduce regulatory friction that slows down AI capacity expansion.
- Why it matters now: The AI industry relies on a relentless, massive scale-up of data centers and GPUs; any delay in permitting or grid access directly impacts the timeline for training next-generation foundational models.
- The underlying tension: Investors are already questioning the immediate return on investment for immense AI infrastructure bets, making any political delay a catalyst for market volatility.
- The infrastructure reality: The bottleneck is no longer just securing silicon, but rather navigating the physical and political realities of power procurement, land use, and regulatory compliance.
What happened
Speaking at the UBS Australasia Conference in Sydney, UBS CEO Sergio Ermotti highlighted a growing risk to the AI sector: political friction. In a Bloomberg Television interview, Ermotti pointed out that the US midterm elections inject a new layer of uncertainty into the technology market. This political shift, coupled with the potential for heightened regulatory oversight, is expected to stretch out the physical construction of artificial intelligence capacity.
Ermotti’s core argument is that tighter regulatory scrutiny and shifting policy priorities will stretch infrastructure project timelines. While he stressed that AI remains a powerful medium- to long-term growth engine, he noted that the short-term picture is changing. The market now faces a reassessment as investors recognize that the timing of economic benefits from AI spending could slip because of political and physical roadblocks.
Why it matters now
The race to build artificial general intelligence (AGI) is fundamentally an infrastructure race. Companies like OpenAI, Google, Anthropic, and Meta rely on continuous, uninterrupted deployments of massive GPU clusters to train their increasingly parameter-heavy LLMs. If the “AI buildout”—a complex web of data centers, cooling systems, high-speed networking, and gigawatt-level power grids—slows because of regulatory red tape, the entire roadmap for next-generation model releases could slip.
This political uncertainty arrives at a moment of market vulnerability. Wall Street is already showing anxiety over hyperscaler capital expenditure (capex), questioning whether the immense upfront costs of AI infrastructure will deliver returns quickly enough. Tighter environmental regulations, complex data-center permitting processes, and grid interconnection delays add pressure on AI development.
For chipmakers and data-center suppliers, the immediate threat is not a lack of demand but a lack of deployable capacity. If hyperscalers cannot secure local permits or clean power contracts for new facilities, their pace of hardware purchases will cool, sending ripples through the AI supply chain and testing the high valuations of tech equities.
Who is most affected
A slowdown in the AI buildout creates a massive ripple effect across the entire intelligence supply chain. The most immediate impact will be felt by the physical infrastructure developers and the silicon designers, but the downstream effects will hit enterprise AI buyers who depend on expanding compute capacity for their own corporate deployments.
| Stakeholder | Impact | Why |
|---|---|---|
| Hyperscalers (Google, Microsoft, AWS) | Delayed capacity scaling | Stricter permitting and grid regulations slow down new data center groundbreakings. |
| Chipmakers (NVIDIA, AMD) | Supply chain bottlenecks | If data centers cannot be built or powered, hyperscalers cannot deploy newly purchased GPUs. |
| Utility Providers & Grid Operators | Heightened scrutiny | Operators must navigate intense political pushback over the massive energy demands of AI facilities. |
| AI Investors | Repricing of risk | Delayed economic benefits force a re-evaluation of current, perfection-priced AI equity valuations. |
About the analysis
Researched and written with AI assistance under the editorial responsibility of the i10X News team. The title image is an AI-generated illustration. Read our AI Use Policy. Spotted an error? Request a correction.
This is an independent, research-based i10x analysis designed for AI founders, infrastructure developers, and technology investors tracking the intersection of policy and compute. It synthesizes executive commentary and global financial reporting to assess the systemic risks to the AI supply chain.
- Bloomberg: UBS CEO Ermotti Says Midterms Create Uncertainty for AI Buildout
- Bloomberg Video: Midterms Creating Uncertainty for AI Buildout
- The Edge Malaysia: UBS CEO Ermotti says midterms create uncertainty for AI buildout
- FourWeekMBA: UBS CEO Ermotti Says Midterms May Slow AI Capacity Buildout
- Business Insider: UBS CEO says US midterm results may slow down AI buildout, Bloomberg says
i10x Perspective
The AI narrative is shifting from a software arms race to a hard-hat reality. While Silicon Valley models scale exponentially, the physical world—governed by elections, grid operators, and local zoning boards—moves at a linear pace. This friction will not end the AI boom, but it will reshape it. The winners of the next decade will be those who can navigate the geopolitical and regulatory maze of building the physical intelligence grid, not only those with the strongest model architectures.
