Industry

Why On-Prem AI Is Back: What Y Combinator, the Data and Regulators Are Saying

On-premises AI means running AI models inside an organisation's own data centre or private cloud, so sensitive data never leaves its control. After years of cloud-first AI, banks and insurers are moving their most sensitive AI work back in-house: the share of enterprises running production AI mainly on public cloud fell from 56% to 41% in a single year. Y Combinator's latest Requests for Startups point the same way.

Singapore · 16 November 2026

Summary

41%

Of enterprises run production AI mainly on public cloud, down from 56% in one year

91%

Would choose on-prem, private or hybrid for AI on sensitive data

5%

Of custom enterprise AI tools reach production

Why is on-prem AI coming back?

For most of the last few years, enterprise AI meant sending data to a model in someone else's cloud. That works for drafting emails. It doesn't work for credit files, medical underwriting records or claims, which most regulated firms can't send outside their own environment.

The numbers show the shift:

FindingNumberSource
Enterprises using public cloud as their main home for production AI inference56% → 41% in one yearBroadcom Private Cloud Outlook 2026, via LLM Configurator
Enterprises that factor an AI vendor's country of origin into selection77%Deloitte State of AI in the Enterprise 2026, via LLM Configurator
Would choose on-prem, private or hybrid for AI on sensitive data91%Cloudian (2026)
Hit new cross-border data restrictions in the past two years45%Cloudian (2026)
Firms with AI repatriation initiatives8% → 16% in a yearBARC Data Sovereignty 2026

The pattern is consistent: sensitive AI work is moving into private infrastructure, while low-risk tasks stay in the cloud.

What is Y Combinator saying about enterprise AI?

YC requestWhat it asks forWhy it points to private AI
Startups that want to sell to huge companies (Summer 2026)Small teams landing Fortune 10 pilots and large deals in year oneBig-company deals pass through security, data-residency and procurement review
AI-native service companies (Summer 2026)AI that delivers finished work in insurance, audit, compliance and healthcare adminEvery one of these sectors is regulated and handles customer data
Company brain (Summer 2026)A structured map of a company's know-how that AI agents can useA company's know-how is its most sensitive data
AI-native compliance infrastructure (Fall 2026)Regulatory monitoring, anomaly flagging and reportingCompliance tools can't leak the data they protect
AI open source and inference infrastructure (Spring 2025)Open-source AI businesses and inference infrastructureOpen models are what make self-hosted AI possible

YC's message to founders is to build AI that does real work inside big, regulated companies. In those companies, the first question is always where the model runs and where the data goes.

What problems do banks and insurers face with AI?

ProblemWhat it looks like
Data sovereigntyRisk and legal teams block sending customer data, credit files or claims to a third-party AI service.
Regulatory accountabilityRegulators expect an AI inventory, testing, human oversight and monitoring, including for AI bought from vendors.
Pilots that never shipDemos impress, then stall on integration, accuracy and security sign-off. MIT research found only 5% of custom enterprise AI tools reach production.
AI that doesn't know the businessGeneral chatbots can't apply the firm's own policies or learn from corrections.
Skills gaps44% of firms cite a lack of staff to run AI infrastructure (BARC).
Unpredictable costUsage-based bills grow quickly once AI agents run every step through a large model.
Shadow AIStaff at over 90% of firms use personal AI tools, but only 40% of firms provide an official one (MIT NANDA).

How are firms dealing with it today?

Most institutions pick one of four routes, and each trades control for effort:

RouteData controlEffort to runWhere it falls short
Cloud AI services with no-training termsContractualLowData still leaves the institution, which fails strict residency reviews
Self-hosting open modelsHighHighNeeds scarce ML and operations staff; a model on its own has no audit trail or workflow
General private AI platformsHighMediumBroad search and chat tools, not the underwriting or claims workflow itself
Custom builds by systems integratorsHighHighSlow and bespoke, often a year or more to production

Many large firms end up hybrid: sensitive steps on private models, low-risk tasks on cloud services. What's usually missing is the last step: an auditable workflow that runs on the institution's own infrastructure and reaches production in weeks.

What does the research say?

  • Small models are enough for most agent work. NVIDIA Research argues in Small Language Models are the Future of Agentic AI that most AI agent tasks are narrow and repetitive. It estimates a 7-billion-parameter model costs 10 to 30 times less to run than a 70 to 175-billion-parameter one.
  • The blocker is workflow fit, not model quality. MIT NANDA's GenAI Divide research found most AI tools fail because they don't learn from feedback or fit how the business works. The projects that succeed target narrow, high-value work in operations and finance.
  • Cost favours on-prem at scale. Above a certain volume, running AI on-prem can save 50% or more over three years compared with usage-based APIs (LLM Configurator).

How does Pints AI solve it?

We've been building private, auditable AI for banks and insurers since 2021, long before on-prem AI became a talking point. Our platform, Autothought, runs AI agents inside the client's own cloud or data centre and links every decision to its source evidence, so it can be replayed for auditors and regulators.

ProblemHow Autothought answers it
Data sovereigntyRuns inside your cloud or on your own servers. Nothing leaves your environment, and you own your data, models and outputs.
Regulatory accountabilityEvery decision is linked to its source documents and rules, with human review built in and a replayable audit trail.
Pilots that never shipPre-built applications for document review, research, onboarding and reporting, and go-live in as little as 21 days.
AI that doesn't know the businessAgents apply your own business rules and policies, not generic ones.
Skills gapsOur team deploys, monitors and supports the platform with you, so you don't need to build an AI operations team first.
Unpredictable costEach task is routed to the right model, including our own small language models, which is how clients have cut token costs by 83%.

The results come from production, not pilots:

MeasureResult
Agent runs completed154 million
Live institutional deploymentsMore than 12
Time to an underwriting decision90% less
Token cost83% lower

Our small-model research is public too. In 2024 we open-sourced 1.5-Pints, a compact language model trained in 9 days that outperformed Apple's OpenELM and Microsoft's Phi-1.5 on the MT-Bench benchmark.

What do MAS’s new AI guidelines mean for on-prem AI?

MAS issued its Guidelines on AI Risk Management on 7 October 2026. The first expectations apply from 7 October 2027, and AI life cycle controls such as human oversight, testing and monitoring follow by 7 October 2028. Institutions remain responsible for AI they buy from vendors.

That makes evidence the deciding factor. Autothought produces it by default, inside your own environment. For the full requirements and deadlines, read our explainer, MAS AI Risk Management Guidelines Explained.

Experience Autothought through co-funding of up to S$200,000 from MAS under the FSTI 4.0 AI Pathfinder track (only for FIs onboarded to MAS’s PathFin.ai programme).

FAQ

What is on-premises AI?

On-premises AI runs AI models inside an organisation's own data centre or private cloud, so sensitive data is processed without leaving its control.

Why are banks moving AI back on-prem?

Data residency rules, regulatory accountability and vendor risk. Broadcom found the share of enterprises running production AI mainly on public cloud fell from 56% to 41% in a year.

Is on-prem AI more expensive than cloud AI?

Upfront costs are higher, but at high volumes on-prem can be cheaper. Routing tasks to smaller models also cuts running costs; Pints AI clients report 83% lower token costs.

Does Autothought run on-premises?

Yes. Autothought runs inside your own cloud or on your own servers, and you keep ownership of your data, models and outputs.

Talk to Pints AI

See how Autothought would work in your environment, with decisions your risk and compliance teams can stand behind.

References

Pints AI performance figures (agent runs, deployments, underwriting time and token cost) are as reported by Pints AI.

Published 16 November 2026. Figures as reported by Pints AI at the date of publication.