Blog Details

The Enterprise Development Grant (EDG) can support up to 50% of qualifying costs for eligible Singapore SMEs running an AI project, covering third-party consultancy, software or equipment, and in some cases internal manpower. It does not fund work retroactively: you need approval before the project starts, and payouts are made on a reimbursement basis after you submit proof of spend. This article walks through eligibility, what actually qualifies, and how the process runs from application to claim.

EDG sits under Enterprise Singapore (EnterpriseSG) and is designed to help local companies upgrade capabilities, innovate, or grow overseas. AI adoption projects, when framed correctly, typically fall under the "Innovation and Productivity" pillar. That framing matters: EDG is not a blanket "AI grant," it is a capability-development grant that AI projects can qualify under when the project is structured as a genuine upgrade to how the business operates.

What is the EDG and how does it apply to AI?

EDG helps Singapore companies defray part of the cost of projects that improve business capability, whether that is process redesign, technology adoption, or market access. An AI project qualifies when it is framed as a capability upgrade, for example automating a manual workflow with a conversational AI agent, or introducing an AI-driven system to replace a slow, error-prone process. A vague request to "add AI" to the business, without a clear before-and-after capability change, is a weaker application.

Who can apply

• The applicant must be a business entity registered and operating in Singapore
• The company must have a minimum of 30% local shareholding, per prevailing EnterpriseSG criteria
• The company should be in a financially viable position to start and complete the project
• Eligibility criteria are set and updated by EnterpriseSG, so always confirm current requirements on the GoBusiness Grants portal or with EnterpriseSG directly before applying

Because eligibility rules are reviewed periodically, treat any secondhand summary (including this one) as a starting point, not the final word. The GoBusiness Grants portal is the authoritative source at the time you apply.

What costs typically qualify under EDG for an AI project?

EDG can support up to 50% of qualifying costs for eligible SMEs, though the exact support level depends on the project and prevailing scheme parameters at the time of application. Costs that are commonly considered qualifying for an AI project include:

• Third-party consultancy fees: the cost of an external partner (like an AI implementation agency) designing, building, and deploying the solution
• Software and equipment: licensing costs, cloud infrastructure tied directly to the project, or hardware required for the solution to function
• Internal manpower: in some project structures, the cost of your own staff time dedicated to the project may be considered, subject to EnterpriseSG's rules on qualifying manpower costs

What is generally NOT considered a qualifying cost: routine operating expenses, work already completed before grant approval, and costs unrelated to the specific capability being built. If you are unsure whether a specific cost qualifies, that is a question for EnterpriseSG or your appointed consultant before you submit, not after.

What does the EDG application process look like?

The general shape of an EDG application, at a high level, runs like this:

1. Scope the project. Define what capability you are upgrading, what the AI solution will actually do, and what the measurable outcome looks like (time saved, error rate reduced, capacity freed up).
2. Engage a vendor or consultant (if using one). Get a proper proposal and costing from your implementation partner, since this typically forms part of the supporting documentation.
3. Submit the application via the GoBusiness Grants portal before any project work begins. This is the step most businesses get wrong: starting work, then applying, generally disqualifies the project from support.
4. Wait for assessment and approval. EnterpriseSG reviews the application. We do not publish or guess at processing times here, since they vary and are set by EnterpriseSG, not by any vendor. Check the GoBusiness portal for current guidance.
5. Execute the project once approved, keeping proper documentation (invoices, contracts, proof of payment, deliverables) throughout.
6. Submit the claim with evidence of completed work and actual spend. The grant is disbursed on a reimbursement basis, meaning you pay first and claim back the supported portion afterwards.

Every application is submitted and assessed in the applicant company's own name. Neither Freemansland nor any other implementation partner applies on your behalf, guarantees approval, or influences EnterpriseSG's decision. Our role, when engaged, is to help scope the AI project properly, provide the technical proposal and costing an application needs, and deliver the actual implementation once approved.

What makes an AI project a strong EDG candidate?

From what we see helping SMEs plan AI adoption, applications tend to be stronger when they:

• Target a specific, well-defined business process rather than a vague "we want AI" ambition
• Have a clear baseline (how the process works today) and a clear target state (how it will work after)
• Come with a costed proposal from a vendor who can explain, in plain terms, what will be built and why
• Show the company has the operational readiness to actually use and sustain the solution after go-live, not just deploy it and let it sit idle

This is the same discipline we walk through in our 90-day AI implementation roadmap article: a strong grant application and a strong implementation plan are usually the same document viewed from two angles. If you have not yet mapped which part of your business would benefit most, our AI opportunity and ROI mapping service is designed to produce exactly that groundwork before you approach a grant application, and you can request a quote at any point in that process if you want a second opinion on scope.

Common misconceptions about EDG and AI

Misconception	Reality
"EDG guarantees funding if I have a good idea"	Approval is assessed by EnterpriseSG against their criteria. No vendor can guarantee an outcome.
"I can start the project and apply for EDG later"	Generally not accepted. Apply and get approval before project work starts.
"EDG pays for the project upfront"	EDG operates on a reimbursement basis. You fund the project, then claim the supported portion back with evidence of spend.
"Any software purchase counts as an AI project"	The project needs to demonstrate a genuine capability upgrade, not just a purchase.

What does EnterpriseSG actually look for in an AI project application?

Assessment criteria are set by EnterpriseSG and can be reviewed over time, so this is not a checklist that guarantees anything. But based on how the scheme is structured around capability development, a few themes come up consistently in how AI projects are framed for EDG:

• Is the project a genuine capability upgrade, or a repackaged purchase? A project that clearly changes how a core business process works tends to demonstrate the "capability" angle EDG is built around more convincingly than a straightforward software purchase dressed up as an AI initiative.
• Is the business ready to execute it? This includes having the operational capacity, the internal buy-in, and (where relevant) the data foundations to actually run the solution once built. A project that looks good on paper but has no realistic path to being used day to day is a weaker candidate.
• Is the scope specific and costed? Vague ambitions ("we want to use AI more") are harder to assess than a defined project with a clear deliverable, timeline, and budget.
• Does the applicant company meet the baseline eligibility criteria? Registration in Singapore, local shareholding thresholds, and financial viability are foundational, and worth confirming early rather than discovering a mismatch after significant scoping work.

None of this is a substitute for reading the current guidance on the GoBusiness Grants portal, which reflects whatever EnterpriseSG's live criteria are at the time you apply.

How does an AI project differ from a typical EDG-funded IT project?

EDG has long supported general capability-building and digitalisation projects, so it is worth being clear about what makes an AI project specifically different from, say, implementing a new accounting system or CRM.

Typical digitalisation project	AI project
Deploys a defined system with largely fixed behaviour	Often involves a system that learns from data, improves with tuning, or handles unstructured input (like natural language)
Success is mostly about correct configuration	Success also depends on data quality, ongoing monitoring, and iteration after go-live
Training is usually a one-time onboarding session	Staff may need to understand how to work alongside an AI system, including when to trust it and when to override it
The "end state" is largely static	The system may need occasional retraining, prompt refinement, or performance monitoring to stay useful

This distinction matters for how you scope the EDG application. A strong AI project proposal should acknowledge that the work does not fully stop at go-live: there is a period of monitoring and refinement afterwards. Our article on what to track after an AI system launches covers what this ongoing discipline looks like in practice, and it is worth building that expectation into your project plan from the start rather than treating go-live as the finish line.

What happens after the project is approved but before you claim?

The period between approval and claim submission is where a lot of the practical discipline lives. During this phase:

• Keep every invoice, contract, and payment record tied to the approved project scope, not mixed in with unrelated business expenses
• Track deliverables against what was proposed, since the claim will need to demonstrate the project was actually completed as scoped
• Flag any material change in scope to EnterpriseSG rather than quietly deviating from the approved project, since significant changes may affect what is claimable
• Keep your implementation partner's documentation (technical specifications, delivery sign-offs) organised alongside your financial records

Businesses that treat this phase with the same discipline as the application itself tend to have a smoother claim process. Businesses that treat approval as the finish line often find themselves scrambling to reconstruct documentation months later.

How this fits into a wider AI strategy

A grant is a financing mechanism, not a strategy. Before applying, it is worth stepping back and asking what problem you are actually solving and whether AI is the right tool for it. Our AI strategy and advisory work exists for that reason: to make sure the project you eventually take to EnterpriseSG is one that will genuinely move the needle for your business, not just one that ticks a grant checkbox. For the funding landscape more broadly, our earlier piece on how Singapore SMEs can fund AI adoption covers PSG and other options alongside EDG.

Ready to see what AI can do for your business?

If you are exploring an AI project and want a clear-eyed view of whether it is EDG-ready, request a quote and we will walk through the scoping with you. You can also reach us directly on WhatsApp at +65 9184 9908 or email glenn@freemansland.co, or head to our contact page to arrange a call. We will always tell you honestly whether a project looks like a strong grant candidate before you spend time on an application.

Frequently Asked Questions