Artificial intelligence has become one of the dominant themes in early-stage company formation over the past three years. The combination of widely available large language model APIs, falling compute costs, and genuine commercial applications across every sector has produced a wave of AI-enabled businesses — many of which are seeking SEIS investment at the seed stage.
This guide covers how SEIS applies to AI companies, what types of AI businesses typically qualify, and what investors should consider when evaluating AI SEIS opportunities.
Do AI companies qualify for SEIS?
The SEIS rules do not have sector-specific exclusions for AI. Whether an AI company qualifies depends on the same factors that apply to any company seeking SEIS: age (less than three years from first commercial sale), size (fewer than 25 employees, assets under £350,000), total raise (under £250,000 through SEIS), and trade (must be a qualifying commercial activity).
Most AI application businesses — software tools, vertical AI products, AI-enabled services — qualify without difficulty. AI infrastructure businesses (model training, compute optimisation) also generally qualify. The exclusions that most often catch AI companies are: financial services (AI trading tools, credit models used by regulated entities), and any activities that HMRC classifies as excluded trades.
The AI SEIS opportunity in 2026
The volume of AI-related SEIS fundraises in the UK has increased substantially over the past two years. According to data from Beauhurst, AI and machine learning companies have been consistently among the most active sectors for seed-stage UK fundraising. Many are raising their first institutional round under SEIS before progressing to EIS as they scale.
The sectors generating the most activity include: AI tools for professional services (legal, accounting, recruitment), vertical AI applications in healthcare and life sciences, AI-enabled B2B software, computer vision applications in manufacturing and logistics, and AI infrastructure tooling for enterprise deployment.
What to look for in an AI SEIS investment
AI is a crowded space and many AI-enabled businesses will not achieve the escape velocity needed to deliver returns at the scale that justifies SEIS risk. Key questions to consider:
- Is the AI genuinely proprietary? Many AI products are thin wrappers around foundation models that can be replicated quickly. Proprietary training data, fine-tuned models, or AI embedded in a workflow with genuine switching costs are more defensible.
- What is the customer retention evidence? Annual contract values and net revenue retention are better early indicators of product-market fit than user acquisition metrics.
- Is the team technical? AI product quality correlates strongly with the depth of technical expertise in the founding team. Domain expertise plus AI capability is the combination that typically produces defensible products.
- What happens when foundation model prices fall further? If the product economics depend on current API pricing, model what happens as compute costs continue to fall.
Also Read
What is SEIS? The complete guide →