The Series A Landscape for AI Companies
The Series A landscape for AI companies in Singapore has evolved significantly. Median Series A round sizes for AI companies in Singapore reached $12M in the second half of 2025, up from $8M in 2023. Pre-money valuations for AI companies with demonstrated product-market fit and $1M+ ARR typically range from $40M to $80M, depending on the sub-sector, growth rate, and competitive dynamics. The investor base has also expanded — while traditional VC funds (Sequoia, Vertex, Jungle Ventures) remain active, we are seeing increased participation from corporate venture arms (particularly from Japanese and Korean conglomerates) and crossover PE funds deploying earlier-stage capital.
This evolving landscape creates both opportunities and challenges for AI founders. More capital and more investors mean more options, but also more complexity in structuring a round that balances founder control, investor protections, and the company's long-term interests. The legal structure of your Series A sets the foundation for every subsequent financing round and, ultimately, for your exit. Getting it right from the start is one of the most impactful things a founder can do.
Preferred Share Structure
Series A investors in Singapore typically receive convertible preferred shares with a package of economic and governance rights. The core economic terms include a liquidation preference (typically 1x non-participating, meaning the investor receives the greater of their investment amount or their pro-rata share of the proceeds on a liquidation event), an anti-dilution protection (typically broad-based weighted average, which adjusts the conversion ratio if the company issues shares at a lower price in a subsequent round), and pro-rata participation rights (the right to invest their pro-rata share in future rounds).
For AI companies, two economic terms deserve particular attention. First, the liquidation preference structure. In a strong fundraising environment, founders should resist participating preferred structures, which allow investors to receive their liquidation preference AND their pro-rata share of remaining proceeds. Participating preferred significantly reduces the founder's payout in moderate exit scenarios. We always model the waterfall analysis for multiple exit scenarios to help founders understand the real-world implications of different preference structures.
Second, the anti-dilution mechanism. The standard broad-based weighted average anti-dilution is founder-friendly relative to full ratchet anti-dilution (which adjusts the conversion price to the lower price without any weighting). However, the specific formula matters — particularly the definition of the denominator in the weighted average calculation. Some investor-drafted formulas exclude certain categories of shares from the denominator, which amplifies the dilution protection. We recommend that founders and their counsel review the anti-dilution formula carefully and model its impact under various down-round scenarios.
Governance & Control
Series A governance terms typically include board composition, protective provisions (veto rights), information rights, and drag-along/tag-along rights. For a Singapore private company, the board is the primary governance body, and its composition directly determines operational control. The standard Series A board structure is 2-1 (two founder-appointed directors and one investor-appointed director) or 2-1-1 (two founder, one investor, one independent). We strongly recommend that AI founders push for the 2-1-1 structure, with the independent director mutually agreed upon, as this provides a natural tiebreaker and demonstrates governance maturity.
Protective provisions — the actions that require investor consent — are among the most negotiated terms in a Series A. Standard protective provisions include approval of future financing rounds, changes to the company's constitution, declaration of dividends, related party transactions, and changes to the size of the board. For AI companies, we recommend that founders pay particular attention to provisions that could restrict the company's ability to pivot its technology, enter new markets, or make small acquisitions. Overly broad protective provisions can effectively give the Series A investor a veto over operational decisions that should remain with management.
AI-Specific Considerations
Several legal considerations are specific to AI companies raising Series A. First, IP representations and warranties. Series A investors will require the company to represent that it owns or has the right to use all IP necessary for its business. For AI companies, this representation covers not only traditional software IP but also trained models, training data, and the algorithms and techniques used in model development. Founders should ensure that these representations are accurate — which means completing an IP audit before the fundraise, not during it.
Second, data rights and regulatory compliance. Investors will want to understand the company's data governance framework, its compliance with the PDPA and any other applicable data protection laws, and its approach to AI governance. Companies that can demonstrate compliance with the Model AI Governance Framework and that have completed AI Verify testing will find the due diligence process significantly smoother. Third, ESOP allocation. The Series A is typically where the ESOP pool is formalized and expanded. Standard ESOP pools at Series A range from 10-15% of the fully diluted share capital, allocated from the pre-money capitalization (which means the dilution falls on the existing shareholders, not the incoming investors). For AI companies, where talent retention is critical, we sometimes negotiate a larger ESOP pool (15-20%) with a structured vesting schedule that balances investor concerns about dilution with the company's need to attract and retain top AI talent.