A value chain under strain

Entangled interests and growing dominance at the top of the market are weakening the AI value chain. 

72% of leaders say their network is increasingly interconnected and complex. Multiple suppliers are owned by the same parent company, and suppliers act simultaneously as investors and collaborators across the ecosystem, raising contagion risk.

Leaders also notice the same providers appearing across their value chain, indicating a shortage of suppliers with the capability and track record required to deliver AI projects.

Increased risks of disruption

Reliance on single suppliers for critical AI operations, particularly in the US, ramps up the risk of disruption further. 49% of companies across the AI ecosystem depend on one cloud provider and 42% have a single source for middleware and APIs. As a result, 67% agree that they are exposed to dependency risks in the AI value chain. 

To regain control, large LLM providers are looking upstream, buying up data centers and chip manufacturing capabilities. While integration deals can boost the resilience of individual supply chains, they’ll also increase concentration risk in the wider AI ecosystem. The cycle begins again.

Data and cloud sovereignty risks

Data is where geopolitics, regulation and commercial strategy collide. As AI becomes central to national security and economic strength, the push for sovereignty adds another dimension of complexity.

39% of leaders believe data and cloud sovereignty will be the biggest test of AI resilience in 2026, making it a top-three disruptor. 29% say tariffs and trade policy will also cause disruption, which is a rise of 10% on the previous year.

There is growing divergence between the US, Europe and the rest of the world on data obligations, limitations and policy. From use rights, consent and regulatory approval, to managing regulated information and moving data across international borders, these differences are making it more difficult to govern data globally.

Cybersecurity threats

As a result, stronger regional value chains will emerge, with domestic AI suppliers and digital infrastructure built to comply with local law. But duplication of AI tools multiplies the cybersecurity threat, which is already a top disruptor for 52% of innovators, adopters and investors.

Liability risks

As organizations integrate AI deeper into core operations and product portfolios, the stakes for liability and safety couldn’t be higher. 

Organizations are trying to shift the burden, limiting their exposure through disclaimers and contractual liability clauses. 65% of leaders say liability is being passed along the value chain, and are unclear what recourse they may have if AI fails.

This strategy could have the opposite effect. Courts won’t tolerate an outcome where no one is accountable. Outside the US, there are new laws reversing burdens of causation and fault. US courts are also looking to traditional tort remedies for multi-defendant claims. 

Shared responsibility

To counter risk on the front end, improving product safety is a top priority for AI innovators. Testing tools in advance means companies can show their record of compliance and due care – and stop problems before they start. 

But AI safety and accuracy is not the sole responsibility of developers. It can fail at several stages. Adopters must also understand how the tools work so they can deploy them properly. While early litigation is currently focused on tech, risk will likely travel downstream to how AI is used by adopters.

Market consolidation

Complexity of AI regulation and cost of compliance is pushing the market toward consolidation, reinforcing the power of large players. 

70% of leaders believe growing regulatory obligations are driving consolidation between AI companies as large players dominate the top of the market. Innovators are incentivized to get hyperscalers and corporate venturing arms involved early to secure the firepower they need to manage legal exposure. 

But this creates dependency when startups align their products and services closely with investor interests, rather than being led by innovation. 71% of leaders believe big cloud providers act as gatekeepers of AI innovation.

Regulatory uncertainty

Despite these concerns, there’s no consensus between innovators, investors and adopters on what an ideal regulatory standard looks like. Leaders recognize the difficulty of striking the right balance between innovation and protection, especially when technology is changing so quickly. But a majority believe the EU AI Act will hurt Europe’s competitiveness in the AI field.

Rising litigation risk in AI deployment

AI at work means double the disputes risk for innovators, investors and adopters. Both employer and employee AI use is proving contentious. 

27% of leaders predict a major dispute or class action in 2026 related to how organizations deploy AI. There are several potential areas of risk, including restructuring and job displacement, appropriate human supervision of AI tools, and rights violations.

Union scrutiny of AI is also expected to reach a turning point this year, according to 26% of organizations. Unions are increasingly sophisticated negotiators on AI, calling for transparency and information sharing. We've already seen companies fined in Europe for failing to meet these obligations.

Policy gaps

At the same time, employee AI misuse is a growing driver of investigations, disciplinaries and disputes. And the policies designed to manage how people use AI may actually expose organizations to litigation risk. 70% of leaders believe their company needs to prepare for a wave of unfair dismissal cases arising from unclear, outdated and ill-conceived AI use policies. 

Risk of intellectual property theft

IP is suffering in the drive for open AI.

64% of innovators agree that explainability and model transparency is exposing them to IP theft – and 40% also see IP and copyright infringement claims as one of their biggest risks in 2026.

But regulatory direction and continued pressure from customers mean transparency is inevitable. 47% of innovators will focus on improvements to explainability and transparency in the year ahead, compared to 20% who plan to prioritize IP enforcement and patent protection. This indicates innovators prefer to get on the front foot to manage risk, rather than relying on litigation, which can be expensive, slow and reputationally sensitive.

The data suggests that existing protection strategies have not fully kept pace with the evolving regulatory and patent environment. As greater transparency around model development is contemplated, maintaining trade secret protection may become more challenging. In that context, some companies are beginning to reassess the role of patents in protecting AI-related software and models.

Mounting copyright pressure

At the same time, developers are defending multiple copyright claims themselves, draining resources and appetite to pursue enforcement litigation. In the US alone, there are currently around 80 active copyright cases against AI developers, creating a heightened state of controversy and driving licensing deals.

Risks of liquidity strain

Investor priorities for 2026 point to financial discipline, with capital deployed more selectively and liquidity strain pushed downstream. 

43% of AI investors will risk-manage portfolio growth in the coming year, investing in targets with a path to profitability, good margins, an ability to run on cheaper hardware and clear data ownership.

This tighter focus means securing working capital on good terms is a struggle for developers, despite rising investment levels. 72% say the cost of capital is going up because of AI bubble fears. 67% believe this will increase reliance on alternative funding from sovereign wealth, family offices and other unregulated capital providers.