Venture capitalists are on a tear to create artificial intelligence’s playoff bracket, piling cash into young startups in the race to harvest viable products and profit. The strategy, long the domain of late-stage wars, is increasingly coming to Series A and B rounds, redefining competitive landscapes for AI applications ranging from enterprise resource planning to IT service management.
Early bets on AI startups get bigger and faster
Here’s DualEntry, an AI ERP challenger that raised a $90 million Series A led by Lightspeed and Khosla Ventures at about a $415 million post-money valuation. Rivals Rillet and Campfire AI both promptly answered the bell with dual rounds from firms including a16z, Iconiq, Sequoia and Accel. The takeaway is clear: pick a horse early, load it with capital and bully the market into treating it as the default.
The conduct is not new, veteran investors say, but the timing is. Firms aren’t waiting until Series C or D to get traction; they are front-loading capital at product formation so that they can have higher ownership before the great growth SOPs. That move reflects a power-law lesson from the last decade that many have learned the hard way: if a category goes exponential, early ownership trumps almost any entry price.
The AI cycle exacerbates the logic behind that. Compute is costly, talent is scarce and you also have the need for strong security and compliance, all of which adds up to actual capital needs even early in a product. In spaces where switching costs compound with data integrations and workflow depth, early financial muscle can translate into lasting moats.
How kingmaking works with AI applications in enterprise
The playbook is simple: fund a possible category winner with enough capital to dominate recruiting, spend and enterprise go-to-market; leverage the signal of elite backers and runway to win early lighthouse customers. In markets with a high degree of vendor procurement, such as ERP, buyers like the sound of a lot of staying power is introduced — the “balance sheet as feature” effect.
That dynamic at play has already reared its head in legal tech, where well-financed AI suppliers like Harvey have shored up giant law firms via a combination of technical capacity and reputational longevity. In ERP, DualEntry’s pitch is a twist on that logic: take the place of complex manual workflows and feed back predictive insights while assuring CFOs and CIOs that Oracle will be around for a long time.
And investors are compressing fundraising timelines, too. (Jaya Gupta of Foundation Capital recently remarked that in hot AI categories, Series Bs are now closing 27–60 days after Series As, a lot of the time with next to no net-new operating data.) The speed isn’t so much about metrics but about denying oxygen to upstarts while customer references are still pliable.
Signals and data behind the surge in AI funding
Market trackers confirm the trend. PitchBook and CB Insights saw a wave of AI “mega-rounds” in 2024, with deal sizes and pre-money valuations growing even as overall venture deal counts softened. Median AI Series A rounds increased substantially from the year before and an increasing percentage of AI financings were concentrated at $50 million or above.
At the same time, alliances by hyperscalers tilt the playing field. Credits and co-selling with cloud providers — and, in some cases, strategic checks — can subsidize training and inference costs. That makes it reasonable for VCs to dump money in early so their portfolio companies can nab expensive GPU time, cement distribution deals, and finance compliance programs that shorten enterprise sales cycles.
Investors also cite the strategic value of category narratives. Once one startup in a market wins the first Fortune 500 references, starts appearing on panels at industry conferences or earns that early SOC 2 or ISO certification, late entrants face a more uphill battle. Kingmaking tries to telescope that window so narrative momentum coagulates into de facto standardization.
The risks of overcapitalization in AI startups
History still counsels caution. Delivery logistics wannabe Convoy and electric scooter company Bird both raised boatloads of cash before crashing under the weight of their operational and market dynamics. Despite which, capital can help hasten product-market fit — but it cannot create it out of thin air. In AI, overfunding can lure teams into sprawl-like roadmaps, high burn profiles and expensive GTM experiments that outstrip learning.
There is also the trap of imitation. If three approximately similar companies all raise nine figures of war chest capital to support their businesses, customer acquisition cost can fly and vendors will end up subsidizing pilots, promising custom features and stashing compute off the P&L sheet through cloud monetary incentives. This can lead to consolidation or acquihires with little return for the late-stage entrants.
What to watch next as AI kingmaking accelerates
Three indicators will show whether early-stage kingmaking leads to winners who stick.
- Depth over breadth: leaders should demonstrate growing net revenue retention and module attach, rather than just logo land grabs.
- Demonstrable unit economics: marginal costs must come down as model architectures and inference routing get better.
- Enterprise trust: security certifications, audit trails and data residency controls must come before procurement cycles.
For now, and until we know more, the armament battle is on. Heavyweight firms like Khosla Ventures, Lightspeed, Accel, a16z, Iconiq and Sequoia are making big bets on AI ERP (enterprise resource planning), IT service management and compliance automation in rapid juxtaposition. Unless the power-law era continues, if two or three early anointed winners prevail, today’s prices can all be justified. Otherwise, 2025’s kingmaking might seem like overreach. Either way, the next phase of AI adoption will be defined as much by checkbooks as by code.
