VCs Predict 2026 as Turning Point for Enterprise AI

Three years after OpenAI's ChatGPT ignited a surge of innovation and attention around artificial intelligence, optimists have consistently heralded AI's imminent integration into enterprise software. This enthusiasm has fueled a boom in enterprise AI startups, backed by substantial investment.

However, many businesses are still struggling to realize tangible benefits from these new AI tools. An MIT survey in August revealed that a staggering 95% of enterprises were not seeing a meaningful return on their AI investments.

So, when will companies truly start to leverage AI for real value? TechCrunch recently surveyed 24 enterprise-focused venture capitalists (VCs), and their consensus points to 2026 as the pivotal year when enterprises will meaningfully adopt AI, derive significant value, and subsequently increase their technology budgets. This prediction, however, echoes similar sentiments from VCs over the past three years, raising the question: will 2026 truly be different?

Anticipated Enterprise AI Trends for 2026

VCs shared their outlook on the trends poised to dominate the enterprise AI landscape:

• Beyond Silver Bullets: Kirby Winfield of Ascend notes that enterprises are realizing large language models (LLMs) aren't universal solutions. The focus will shift to custom models, fine-tuning, evaluations, observability, orchestration, and data sovereignty.
• Product to Consulting Shift: Molly Alter from Northzone predicts a subset of enterprise AI companies will transition from product-centric businesses to AI consulting. Starting with specific products like AI customer support, they will leverage customer workflows to build additional use cases, becoming generalist AI implementers.
• Voice AI's Rise: Marcie Vu of Greycroft expresses excitement for voice AI, seeing it as a more natural and efficient interaction method with machines, moving beyond screen-based interfaces.
• AI in the Physical World: Alexa von Tobel of Inspired Capital believes 2026 will see AI reshape physical domains, particularly infrastructure, manufacturing, and climate monitoring, moving from reactive to predictive systems.
• Frontier Labs' Direct Applications: Lonne Jaffe from Insight Partners observes frontier labs potentially shipping more turnkey AI applications directly into production for sectors like finance, law, healthcare, and education, rather than just training models.
• Quantum Momentum: Tom Henriksson of OpenOcean highlights "momentum" for quantum computing in 2026, with increasing trust and roadmaps. However, major software breakthroughs are still awaiting further hardware performance.

Key Investment Areas for VCs

Investors are strategically positioning their capital in several critical areas:

• Physical AI and Model Research: Emily Zhao of Salesforce Ventures targets AI's integration into the physical world and the next evolution of model research.
• Future Datacenter Technology: Michael Stewart from M12 is investing in "token factory" technology, focusing on efficiency and cleanliness within data centers, including cooling, compute, memory, and networking.
• Vertical Enterprise Software: Jonathan Lehr of Work-Bench seeks vertical enterprise software where proprietary workflows and data create defensibility, especially in regulated industries, supply chain, and retail.
• Energy Efficiency for AI: Aaron Jacobson of NEA is looking for software and hardware innovations that improve performance per watt, addressing the energy demands of GPUs through better management, efficient chips, or next-gen networking.

Defining an AI Startup's "Moat"

When evaluating AI startups, VCs emphasize defensibility beyond just model performance:

• Economics and Integration: Rob Biederman of Asymmetric Capital Partners states that a moat is less about the model and more about economics and deep integration into enterprise workflows, proprietary data access, and high switching costs.
• Beyond Model Performance: Jake Flomenberg of Wing Venture Capital is skeptical of moats built solely on model performance or prompting, which erode quickly. He asks if a company would still be relevant if a major new model were released.
• Vertical and Workflow Moats: Molly Alter highlights that moats are easier to build in vertical categories. Data moats, where each interaction improves the product, are strong in specialized fields like manufacturing or healthcare. Workflow moats come from deep understanding of industry-specific processes.
• Data Transformation: Harsha Kapre of Snowflake Ventures believes the strongest moat lies in how effectively a startup transforms an enterprise's existing data into better decisions and experiences, bringing domain-specific solutions directly to governed data.

Will Enterprises See Value and Increase AI Budgets in 2026?

The sentiment is cautiously optimistic, with nuances:

• Focused Engagement: Kirby Winfield suggests enterprises will move away from random experiments to fewer, more thoughtful AI solutions.
• AI as a Scapegoat: Antonia Dean of Black Operator Ventures warns that some enterprises might cite AI investment increases to justify cuts elsewhere, potentially using AI as a scapegoat for past mistakes.
• Application Layer Maturation: Scott Beechuk of Norwest Venture Partners believes 2026 will show if the application layer can convert infrastructure investments into real value as specialized models mature.
• Incremental Value: Marell Evans of Exceptional Capital expects incremental value, with AI improving to solve specific pain points and simulation-to-reality training opening new opportunities.
• Value Already Gained: Jennifer Li of Andreessen Horowitz argues that enterprises are already gaining value (e.g., AI coding tools) and this will multiply next year.
• Budget Shifts, Not Just Increases: Rajeev Dham of Sapphire predicts a shift of labor spend towards AI or AI capabilities generating such strong ROI that investments pay for themselves multiple times over.
• Concentrated Spending: Rob Biederman and Gordon Ritter (Emergence Capital) foresee budgets increasing for a narrow set of AI products that deliver clear results, with spending concentrating on solutions that expand institutional advantages rather than just automating workflows.
• Pushback on Vendor Sprawl: Andrew Ferguson of Databricks Ventures expects CIOs to push back on the proliferation of AI vendors, rationalizing overlapping tools and deploying savings into proven AI technologies.
• Transition from Pilots: Ryan Isono of Maverick Ventures anticipates a shift from experimental budgets to dedicated line items, as enterprises realize the complexity of building in-house AI solutions at scale.

Raising a Series A as an Enterprise AI Startup in 2026

For AI startups seeking Series A funding, VCs emphasize a combination of compelling narrative and concrete traction:

• "Why Now" and Enterprise Adoption: Jake Flomenberg looks for a strong "why now" narrative, often tied to GenAI creating new opportunities, coupled with concrete proof of enterprise adoption (e.g., $1M-$2M ARR baseline, customers viewing the product as mission-critical).
• Expanding Total Addressable Market (TAM): Lonne Jaffe advises building in spaces where AI drives down costs but expands the TAM, rather than evaporating it.
• Real-World Customer Usage: Jonathan Lehr stresses that customers must be using the product in day-to-day operations, willing to provide references on impact, reliability, and the buying process. Products should clearly demonstrate time savings, cost reduction, or increased output.
• Customer Interest and Conversions: Michael Stewart notes that investors are now valuing customer interest and willingness to evaluate solutions. Strong marketing and quality are essential to secure engagements, with conversions expected to be a leading part of the story after six months of pilot use.
• Execution and Traction: Marell Evans prioritizes execution, traction, genuinely delighted users, and the technical sophistication of the business, looking for real contractual agreements and the ability to attract top-tier talent.

The Role of AI Agents in Enterprises by End of 2026

The future of AI agents is seen as a gradual but transformative shift:

• Initial Adoption Phase: Nnamdi Okike of 645 Ventures believes agents will still be in their initial adoption phase, requiring technical, compliance, and communication standards to be overcome.
• Emergence of Universal Agents: Rajeev Dham predicts the emergence of a universal agent, converging siloed roles (e.g., sales, customer support) into a single agent with shared context and memory, breaking down organizational silos.
• Collaborative Augmentation: Antonia Dean foresees sophisticated collaboration between humans and agents on complex tasks, with organizations finding the right balance of autonomy and oversight, viewing agents as collaborative augmentation.
• Agentic Co-workers: Aaron Jacobson humorously suggests that most knowledge workers will have at least one agentic co-worker they know by name.
• Dominant Workforce: Eric Bahn of Hustle Fund speculates that AI agents could become a larger part of the workforce than humans in enterprises due to their near-zero marginal cost.

Strongest Growth and Retention in VC Portfolios

VCs highlight specific characteristics of their fastest-growing and highest-retention portfolio companies:

• Growth Drivers: Jake Flomenberg points to companies that identified and relentlessly executed on workflow or security gaps created by GenAI adoption, such as data security tools for LLMs, agent governance, or new marketing areas like Answer Engine Optimization (AEO). Andrew Ferguson sees growth in companies that land with focused use cases, nail them, and then expand. Jennifer Li notes success in companies that help enterprises put AI into production, covering areas like data extraction, developer productivity, and generative media infrastructure.
• Retention Factors: Jake Flomenberg identifies strong retention in companies that solve problems intensifying with increased AI deployment, are mission-critical, accumulate proprietary context, and address growing rather than one-off issues. Tom Henriksson highlights serious enterprise software providers, especially those enhanced with AI, that deeply transform operations and build proprietary data, making them indispensable. Michael Stewart notes high retention in data tooling and vertical AI apps with forward-deployed teams. Jonathan Lehr emphasizes software that becomes foundational infrastructure rather than a point solution, citing examples like authorization systems or orchestration layers for end-to-end workflows.