The Complete Guide to MDEC MDAG-AI Grant for Enterprise AI Projects (2026)

The Malaysian Digital Acceleration Grant for Artificial Intelligence (MDAG-AI) is the single most impactful government incentive available to Malaysian enterprises investing in AI capabilities. Administered by the Malaysia Digital Economy Corporation (MDEC), the programme provides matching grants of up to RM1 million per qualifying project, effectively halving the cost of enterprise AI advisory, implementation, and infrastructure investments. Since its inception, MDAG-AI has disbursed over RM120 million across 200+ approved projects, making it one of the largest AI-specific grant programmes in Southeast Asia.

Despite its scale and generosity, the MDAG-AI programme is underutilised by the companies that would benefit most from it. MDEC's own data indicates that fewer than 35% of eligible mid-market enterprises have applied, and among those that do apply, the first-submission approval rate is approximately 45%. The gap between eligible companies and successful applicants is not a funding constraint — MDAG-AI has not exhausted its annual allocation in any year since launch. The gap is an information and preparation constraint: companies either do not know the programme exists, do not understand the eligibility requirements, or submit applications that do not meet the evaluation criteria.

This guide is designed to close that gap. It covers every aspect of the MDAG-AI application process in the detail required to submit a successful application on the first attempt. The information is current as of April 2026 and reflects the latest programme guidelines, evaluation criteria, and procedural requirements.

What MDAG-AI Covers: Scope and Funding Structure

MDAG-AI operates as a matching grant, which means MDEC funds up to 50% of qualifying project costs, with the applicant company funding the remaining 50% or more. The maximum grant amount is RM1 million per project. A single company may submit multiple applications for distinct AI projects, subject to a cumulative cap of RM2 million across all approved grants.

Qualifying expenditure categories include AI advisory and consulting services (strategy development, use-case identification, readiness assessments), AI solution development and implementation (model development, data engineering, system integration), AI infrastructure and platform costs (cloud computing, data storage, MLOps tooling — limited to the first 12 months of operational costs), AI talent development (training programmes, certification costs, upskilling for existing staff), and data preparation and governance (data quality remediation, data catalogue development, governance framework implementation). Hardware procurement is not covered unless it is directly and exclusively used for AI model training or inference — general-purpose IT infrastructure upgrades do not qualify.

• /Grant type: Matching grant — MDEC covers up to 50% of qualifying costs.
• /Maximum per project: RM1,000,000.
• /Cumulative cap per company: RM2,000,000 across all approved MDAG-AI grants.
• /Qualifying categories: AI advisory, solution development, platform costs (12 months), talent development, data preparation.
• /Non-qualifying: General IT hardware, non-AI software licenses, marketing costs, overhead allocation.
• /Funding mechanism: Reimbursement-based — company pays upfront, claims reimbursement upon milestone completion.

Eligibility Criteria: Who Can Apply

The eligibility requirements for MDAG-AI are designed to target established Malaysian enterprises with genuine AI adoption intent, while excluding speculative applications and companies that lack the organisational foundation to benefit from AI investment.

• /Company registration: Must be incorporated in Malaysia under the Companies Act 2016 with at least 60% Malaysian ownership.
• /Revenue threshold: Minimum annual revenue of RM1 million in the most recent audited financial year. No maximum revenue cap, but companies with revenue exceeding RM500 million are assessed under stricter additionality criteria.
• /Operating history: Minimum 2 years of continuous business operations at the time of application.
• /Tax compliance: Must hold a valid tax clearance certificate from LHDN (Inland Revenue Board) with no outstanding tax obligations.
• /MSC Malaysia status: Not required but preferred — MSC-status companies receive an expedited review timeline.
• /Sector: Open to all sectors. Priority scoring is given to manufacturing, healthcare, agriculture, financial services, and logistics — sectors identified in the National AI Roadmap as strategic AI adoption priorities.
• /Previous grants: Companies with existing MDAG-AI grants may apply for additional projects provided the cumulative cap is not exceeded and previous grant milestones are on track.
• /Exclusions: Government-linked companies (GLCs) with majority government ownership, sole proprietorships, and companies under judicial management are not eligible.

The Application Process: Step by Step

The MDAG-AI application process consists of five stages, from initial registration through to grant disbursement. Understanding the full sequence before starting the application prevents common procedural errors that cause delays or rejections.

Stage 1: Pre-Application Registration (1-2 weeks)

Before submitting a formal application, companies must register on the MDEC Digital Economy Portal and complete a pre-qualification questionnaire. The questionnaire captures basic company information, the proposed AI project scope, estimated budget, and a self-assessed readiness profile. This stage serves as a preliminary filter — MDEC uses the questionnaire responses to identify applications that are clearly ineligible or insufficiently developed, and provides feedback to redirect these companies to appropriate preparatory resources (such as the SME Digitalisation Grant for earlier-stage companies). Companies that pass pre-qualification receive an Application Reference Number and access to the full application portal.

Stage 2: Full Application Submission (3-4 weeks of preparation)

The full application is the most substantive component of the process and the stage where most rejections occur. The application requires seven components, each of which must be complete and internally consistent.

• /Project Proposal: A detailed description of the AI capability to be built, including the business problem it addresses, the AI methodology to be used, the data assets required, and the expected business outcomes. Maximum 15 pages.
• /Implementation Plan: A milestone-based project plan with defined deliverables, timelines, resource allocation, and dependencies. Must include at least three measurable milestones with quantified success criteria.
• /Budget Breakdown: Line-item budget with quotations or cost estimates for each expenditure category. MDEC requires at least two competitive quotations for any single line item exceeding RM50,000.
• /Business Case: Quantified ROI projection with a minimum 3-year forecast, including baseline metrics (current state), projected improvements, and the methodology used to calculate financial impact. Sensitivity analysis is expected for projections exceeding RM2 million in cumulative benefit.
• /Team Composition: Profiles of the internal project team and any external partners, including relevant AI implementation experience. MDEC evaluates whether the team has the capability to execute the proposed project.
• /Data Readiness Statement: A declaration of the data assets available for the project, including data sources, volume, quality assessment, and any data preparation work required before AI model development can begin.
• /Company Financials: Audited financial statements for the two most recent financial years, plus a current management accounts statement not more than 3 months old.

Stage 3: Technical Review (4-6 weeks)

Submitted applications are reviewed by MDEC's AI Advisory Panel, a rotating group of industry practitioners, academics, and MDEC technical staff. The technical review evaluates the application across four weighted criteria: Technical Feasibility (30% weighting) — is the proposed AI approach sound, and does the applicant have the data and infrastructure to execute it? Business Impact (25%) — will the project generate measurable, sustainable business value? Team Capability (25%) — does the project team have the skills and experience to deliver? Strategic Alignment (20%) — does the project align with Malaysia's National AI Roadmap priorities?

Applications scoring above 70% across all criteria are recommended for approval. Applications scoring between 55% and 70% are referred for a Clarification Interview — a 60-minute session where the applicant presents the project to the review panel and responds to technical and commercial questions. Applications scoring below 55% are rejected with written feedback. The clarification interview is not a negative signal — approximately 60% of companies referred to clarification are ultimately approved. However, the interview adds 3 to 4 weeks to the timeline, which is why first-submission quality matters.

Stage 4: Grant Offer and Agreement (2-3 weeks)

Approved applications receive a formal Grant Offer Letter from MDEC, which specifies the approved grant amount, the milestone schedule, the disbursement conditions, and the reporting obligations. The company has 30 days to accept the offer and execute the Grant Agreement. The agreement is a binding legal document that commits both parties: MDEC commits to reimbursing approved costs upon milestone completion, and the company commits to executing the project as proposed, maintaining accurate financial records, and submitting milestone completion reports within 30 days of each milestone date.

Stage 5: Implementation and Disbursement (6-18 months)

Once the Grant Agreement is executed, the company proceeds with project implementation according to the approved milestone schedule. At each milestone, the company submits a Milestone Completion Report including evidence of deliverable completion, actual expenditure documentation (invoices, receipts, bank statements), and a progress narrative describing any variances from the original plan. MDEC reviews each milestone report within 30 working days and, upon approval, disburses the corresponding grant tranche. The final 20% of the grant is withheld until a Post-Completion Review conducted 3 months after the final milestone, which verifies that the AI capability is operational and generating the projected business value.

Timeline: End-to-End from Application to Full Disbursement

• /Pre-application registration: 1-2 weeks.
• /Application preparation: 3-4 weeks (can be done in parallel with registration).
• /Technical review: 4-6 weeks after submission.
• /Clarification interview (if required): Additional 3-4 weeks.
• /Grant offer and agreement: 2-3 weeks after approval.
• /Total time from start to grant agreement: 10-15 weeks (best case) to 16-21 weeks (if clarification required).
• /Implementation period: 6-18 months (defined in approved milestone schedule).
• /Post-completion review: 3 months after final milestone.
• /Final disbursement: 30 working days after post-completion review approval.
• /End-to-end from application start to full disbursement: 12-24 months depending on project scope.

The 7 Most Common Mistakes That Get Applications Rejected

TechShift has supported 23 successful MDAG-AI applications and reviewed over 50 applications from prospective clients. The rejection patterns are remarkably consistent. Avoiding these seven mistakes will place your application in the top quartile of submissions.

Mistake 1: Vague Business Problem Definition

The most common rejection reason is a project proposal that describes the AI technology to be deployed without clearly articulating the specific business problem it solves. "Implement AI for operational efficiency" is not a business problem — it is a technology preference. "Reduce unplanned downtime on CNC machining lines from 140 hours per year to under 40 hours using predictive maintenance" is a business problem with a measurable outcome. The review panel needs to understand the pain point, the current cost of the pain point, and how AI specifically addresses it.

Mistake 2: Missing Baseline Metrics

The business case requires quantified improvement projections, which require a documented baseline. Applications that project "30% improvement in quality" without specifying the current defect rate, the measurement methodology, and the data source for the baseline figure are routinely scored below threshold on the Business Impact criterion. Establish your baselines before writing the application.

Mistake 3: Unrealistic Timelines

Applications that propose deploying enterprise AI capabilities in 4 to 6 weeks raise credibility concerns with the review panel. Conversely, applications with 24-month implementation plans for straightforward use cases suggest the team lacks execution capability. Realistic timelines for most mid-market AI projects fall between 3 and 12 months for initial production deployment, with a 6-month sweet spot for focused, well-scoped projects.

Mistake 4: Insufficient Team Credentials

The Team Capability criterion carries 25% of the evaluation weight. Applications that list team members by title only — without describing their relevant AI implementation experience — score poorly. For projects that rely on external partners, the partner's track record must be documented with the same rigour as the internal team. If your internal team lacks AI implementation experience, this is not disqualifying — but you must demonstrate that the external partner has the experience and that a knowledge transfer plan is in place.

Mistake 5: Budget Misalignment with Scope

Budgets that are too low for the proposed scope suggest the applicant does not understand the true cost of AI implementation. Budgets that are inflated relative to the scope suggest the applicant is optimising for maximum grant capture rather than project success. The review panel is experienced enough to identify both patterns. Ensure every budget line item is justified by the implementation plan and supported by realistic cost estimates.

Mistake 6: Ignoring Data Readiness

Applications that describe sophisticated AI capabilities without addressing data readiness are flagged as technically infeasible. If your data assets require significant preparation before model development can begin, include a data preparation phase in the implementation plan and budget — and be honest about the current state. An application that acknowledges data gaps and includes a plan to address them is far stronger than one that pretends the data is ready when it is not.

Mistake 7: No Post-Grant Sustainability Plan

MDEC explicitly evaluates whether the AI capability will continue to generate value after the grant period ends. Applications that do not describe how the AI system will be maintained, monitored, and evolved after implementation — including the internal team responsible and the ongoing operational budget — are scored down on the Strategic Alignment criterion. The review panel has seen too many grant-funded projects that delivered a working system and then degraded within 12 months because no one was responsible for keeping it operational.

How TechShift Maximises Your MDAG-AI Approval Odds

TechShift's Grant Navigator service is designed specifically to maximise the probability and speed of MDAG-AI approval. The service includes three components that directly address the most common failure points.

• /Project Design for Grant Alignment: We design your AI project with MDEC evaluation criteria built into the architecture from day one. The grant application is a natural output of good project design, not a separate proposal-writing exercise.
• /Baseline Measurement Sprint: A 2-week assessment that establishes the quantified baselines required for a compelling business case — current downtime hours, defect rates, OEE scores, process cycle times, and cost metrics.
• /Application Development and Review: We prepare the full seven-component application package, including the technical proposal, implementation plan, budget breakdown, and business case — all calibrated to the scoring criteria used by MDEC's review panel.
• /First-submission approval rate: 87% across 23 supported applications, vs industry average of approximately 45%.
• /Average time from engagement start to grant agreement execution: 14 weeks.

Stacking MDAG-AI with Other Incentives

MDAG-AI can be combined with other Malaysian government incentives to further reduce the net cost of AI investment. The most valuable combination for manufacturers is MDAG-AI plus GITA (Green Investment Tax Allowance), which provides a 60% tax deduction on qualifying AI and automation capital expenditure. For technology companies, MDAG-AI plus MSC Malaysia tax incentives can reduce the effective tax rate on AI-related revenue to as low as 5% for qualifying activities. The SME Digitalisation Grant (up to RM5,000 per company) can fund initial AI readiness assessments that lay the groundwork for a subsequent MDAG-AI application.

• /MDAG-AI + GITA: 50% matching grant plus 60% tax deduction on qualifying capex. Combined effective cost reduction: 62-72%.
• /MDAG-AI + MSC Malaysia: Matching grant plus reduced corporate tax rate (5-15%) on qualifying digital economy activities.
• /MDAG-AI + SME Digitalisation Grant: Use the RM5,000 SME grant to fund an initial AI readiness assessment, then leverage the assessment outputs to support the MDAG-AI application.
• /MDAG-AI + HRDF (Human Resource Development Fund): Use HRDF claims to fund the AI talent development component, reducing the MDAG-AI budget allocation required for training.
• /Important: Each incentive has its own application process and eligibility criteria. Ensure compliance with all programme requirements to avoid clawback risk.

Next Steps: Start Your Application Today

The MDAG-AI programme represents a significant opportunity for Malaysian enterprises to accelerate their AI adoption at half the cost. The current funding cycle is active with available allocation, but the December 2026 GITA deadline creates urgency for manufacturers who want to stack both incentives. Companies that begin the application process now can expect to have an approved grant agreement in place within 10 to 15 weeks — well before the GITA deadline.

TechShift's Grant Navigator service provides end-to-end support from initial project scoping through to grant agreement execution. Our 87% first-submission approval rate means your application has nearly double the probability of approval compared to self-submitted applications. More importantly, the project design process that underpins our grant approach produces better-structured AI initiatives that deliver measurable value regardless of whether the grant is approved. The grant is the accelerant — the AI capability is the destination.