TL;DR
A TechCrunch survey of 24 enterprise-focused venture capitalists finds most expect 2026 to be the year enterprises begin to meaningfully adopt AI and lift budgets. That optimism comes amid evidence — including a recent MIT survey — that many companies so far are not seeing meaningful returns from AI investments.
What happened
TechCrunch asked 24 VCs who focus on enterprise software about trends and investment priorities for 2026. Respondents repeatedly named several themes: a move away from one-size-fits-all LLM solutions toward custom models, fine-tuning, evaluation tooling, observability and orchestration; growing interest in voice-first interfaces; and AI applied to the physical world such as infrastructure, manufacturing and climate monitoring. Several investors highlighted datacenter- and energy-efficiency plays — better chip management, cooling, networking and next-generation compute — as priorities for backing. On defensibility, many VCs argued moats won’t come from model size alone but from entrenched workflows, proprietary or continually improving data, and integration with governed enterprise data. Some warned budget increases will be selective: spending may grow overall but concentrate on vendors and products that demonstrably deliver results. The survey echoes prior yearly predictions that 2026 will be decisive for enterprise AI adoption.
Why it matters
- If enterprises shift from experimentation to focused deployments, vendors that integrate tightly with workflows and data may capture disproportionate budget share.
- Investment emphasis on datacenter efficiency and energy-per-watt could reshape hardware and operations priorities in AI implementations.
- A move toward custom models, observability and orchestration reflects a demand for trustworthy, auditable AI in regulated and operational settings.
- Broad claims about ROI remain contested: current surveys show many enterprises are not yet seeing meaningful returns, heightening scrutiny of near-term outcomes.
Key facts
- TechCrunch surveyed 24 enterprise-focused venture capitalists about trends and investments for 2026.
- An MIT survey cited in the coverage found 95% of enterprises reported not getting a meaningful return on AI investments.
- VCs expect a shift toward custom models, fine-tuning, evaluation tooling, observability, orchestration, and data sovereignty.
- Several investors predict growth in voice AI and applications that bring AI into the physical world (infrastructure, manufacturing, climate monitoring).
- Frontier labs may ship more turnkey, production-ready applications into regulated domains such as finance, law, healthcare and education.
- Investors called out datacenter and energy-efficiency investments: cooling, compute, memory, networking and chip/per-watt performance.
- Many VCs said durable moats come from proprietary or continuously improving data, workflow integration and switching costs rather than raw model performance.
- Some VCs expect enterprises to increase AI budgets in 2026, but spending may concentrate on a small set of vendors that show clear results.
- The prediction that 2026 will be a breakout year for enterprise AI adoption has been repeated by VCs for multiple years.
What to watch next
- Whether enterprises actually realize meaningful ROI from AI deployments in 2026 — not confirmed in the source.
- If enterprise AI budgets grow and become concentrated among a few vendors delivering clear outcomes — not confirmed in the source.
- The pace at which voice-first interfaces and physical-world AI (infrastructure, manufacturing, climate monitoring) move from pilots to production — not confirmed in the source.
Quick glossary
- Large language model (LLM): A machine learning model trained on large text corpora to generate or analyze human language; used for tasks like summarization, translation and question answering.
- Data moat: A competitive advantage that comes from proprietary, hard-to-replicate data or continuously improving datasets that improve a product over time.
- Observability: Tools and practices that let teams monitor, trace, and understand system behavior and performance, often used to detect and diagnose issues in production.
- Frontier lab: Research-focused organization or team that develops cutting-edge models or foundational AI capabilities, sometimes referred to as model labs.
- Token factory (in datacenter context): A term used by investors to describe systems and infrastructure centered on efficient generation, processing or handling of tokens used by AI models.
Reader FAQ
Did VCs surveyed expect enterprise AI adoption to accelerate in 2026?
Yes. Most respondents in the TechCrunch survey said they expect 2026 to be the year enterprises start to meaningfully adopt AI.
Are enterprises currently seeing returns from AI investments?
Not according to an MIT survey cited in the coverage, which reported 95% of enterprises were not getting a meaningful return on AI investments.
Will enterprise AI budgets increase in 2026?
Many VCs in the survey expect budgets to rise, but several said increases will be nuanced and concentrated on products that deliver clear results.
What kinds of technologies are VCs looking to back?
Investors named custom models, observability and orchestration, voice AI, physical-world applications, and datacenter energy and efficiency technologies.
Are these predictions new?
No. The coverage notes that VCs have been predicting a breakout year for enterprise AI for multiple years; whether 2026 will differ remains to be seen.
It’s been three years since OpenAI released ChatGPT and kicked off a surge in innovation and attention on AI. Since then, optimists have regularly claimed that AI will become a…
Sources
- VCs predict strong enterprise AI adoption next year — again
- 33 predictions for 2026 in AI: consumer, enterprise, AI capex …
- 10 AI Predictions For 2026
- What's ahead for startups and VCs in 2026? Investors …
Related posts
- Samsung to bring Google Photos app to its smart TVs in 2026 — planned launch
- Three new MacBooks expected early 2026: M5 Air, M5 Pro/Max, and 12.9-inch
- Show HN: See what readers who loved your favorite book/author also loved to read