Budget 2026: Insights on Singapore’s AI initiatives
Initiatives across governance, infrastructure, enterprise support, and workforce development.
LEGAL PULSE · SINGAPORE · MAY 2026
Continuing our take on the Budget 2026, in 3 weeks, Singapore published a coordinated set of AI policy documents covering workforce training, enterprise support, physical infrastructure, and sector governance. This is a stacking of commitments across multiple agencies, all pointing in the same direction.
STRATEGIC GOVERNANCE
National AI Council
New inter-ministerial body chaired by PM Lawrence Wong, with DPM Gan Kim Yong, Minister Ong Ye Kung, Minister Josephine Teo, and Dr Tan See Leng. Provides strategic direction and coordinates the national AI agenda across sectors and agencies.
Chairing this at PM level matters signals that AI is being treated as national infrastructure. The sequencing of which sectors move first will be worth watching once the Council is operational.
No dedicated page yet — the council was announced in the Budget Statement and is expected to be formally constituted later in 2026.
Source: Budget Statement, Section C — singaporebudget.gov.sg
National AI Missions (4 sectors)
4 sector-focused AI transformation programs targeting advanced manufacturing, connectivity and logistics, finance, and healthcare. Each mission is anchored in defined objectives and measurable outcomes.
Source: Budget Statement, Section C — singaporebudget.gov.sg
Model AI Governance Framework for Agentic AI
Published in January 2026, this is the world’s first government framework for agentic AI systems. The core principle: AI can automate tasks, but humans remain meaningfully accountable for outcomes.
For lawyers, this is a formalisation of something that should already be clear. Accountability always follows the professional, not the tool. AI-assisted research, document review, or drafting still carries your name. The framework raises the expectation that practitioners understand what they’re deploying, including agentic systems that operate across multiple steps without direct human oversight.
Source: IMDA press release · MDDI announcement · Full framework PDF
INFRASTRUCTURE
Kampong AI Park at One-North
Singapore’s first purpose-built AI park, developed by JTC at LaunchPad @ One-North. Pilot phase from March 2026; completion by 2028. The AI block provides 14,500 sqm of business park space for up to 70 companies, with an adjacent residential block offering over 200 units — a live-work cluster for the AI community.
The Lorong AI co-working space (22 Cross Street and Vidacity) continues to serve enterprises during construction.
Source: JTC press release — jtc.gov.sg · LaunchPad @ One-North
ENTERPRISE SUPPORT
Champions of AI Programme
Jointly led by EnterpriseSG and DISG. End-to-end AI transformation support for selected Singapore companies, covering enterprise strategy and workforce training. Designed to set benchmarks that inspire broader sector adoption.
The selection-based model means not every firm qualifies. For those that do, the end-to-end framing might be more substantive than a standalone grant. Worth checking eligibility if your firm is mid-transformation.
Source: EnterpriseSG media release
Enterprise Innovation Scheme (EIS) expanded
400% tax deduction on qualifying AI expenditure for Years of Assessment 2027–2028, capped at S$50,000 per year. Qualifying spend may include AI software subscriptions, implementation consultancy, and staff AI training, but IRAS will provide further guidance later this year.
The cap is low for larger practices. S$50,000 of qualifying spend nets S$200,000 in deductions, a significant number for boutiques and mid-size firms that have been sitting on the fence about committing to an AI tool stack. For larger firms, this might not move the needle significantly.
Source: IRAS — Enterprise Innovation Scheme · IRAS — Budget 2026 tax changes
Productivity Solutions Grant (PSG) expanded
AI-enabled solutions in the PSG catalogue will be expanded from the present 30% to 50% of pre-approved tools. A new GenAI Navigator for SMEs has been introduced, recommending pre-approved tools matched to specific business needs, with grant support attached.
Pre-approval reduces the research burden for smaller practices without a dedicated technology function. The GenAI Navigator is a useful starting point, although it remains to be seen how the recommendations hold up in practice once the catalogue is live.
Source: IMDA — PSG programme page · NAIIP Factsheet — imda.gov.sg
WORKFORCE DEVELOPMENT
National AI Impact Programme (NAIIP)
The umbrella programme for workforce AI adoption. Targets 100,000 AI Bilingual workers and 10,000 enterprises over the next three years. Legal and accountancy are the first professions in scope under the TeSA expansion.
The legal cohort is being developed with the Singapore Academy of Law and Singapore Corporate Counsel Association. Three competency areas: redesigning workflows with AI, hands-on tool experience, and responsible AI use including bias awareness. Named use cases are research, document review, and contract management. Programme launch set for H1 2026.
Legal being named first of only two professions is the meaningful part. The profession is being treated as a pilot for what AI Bilingual looks like in a regulated services context. 100,000 workers and 10,000 enterprises in three years is an ambitious target; the legal cohort will be a signal of whether the programme can deliver depth together with volume.
Source: MDDI factsheet · IMDA factsheet
SkillsFuture AI learning pathways
Redesigned MySkillsFuture portal with clearer AI course pathways matched to proficiency level. A self-diagnostic AI Readiness Tool, developed with the Singapore Institute of Technology, allows users to assess their readiness and receive personalised course recommendations. Rolling out Q2 2026.
Source: MySkillsFuture — AI courses · COS 2026: companies and workers towards AI literacy
Six months of free premium AI tool access
Singaporeans who complete selected SkillsFuture AI courses will receive six months of free access to premium AI tools. Partners confirmed: Google, Microsoft, OpenAI, and Manus. Applies to courses completed from H2 2026.
The inclusion of Manus alongside more established names suggests the government is watching what practitioners actually use rather than defaulting to the usual list.
No dedicated page yet — expected on the SkillsFuture website from H2 2026 when the rollout begins; announced in the Budget Statement, Section C.
IHL AI literacy strengthening
AI education reinforced across all Institutes of Higher Learning, with emphasis on rigorous disciplinary thinking alongside AI fluency. The aim is graduates with disciplinary depth who can direct AI effectively — not graduates who’ve learned to use a few tools.
Source: MOE — COS 2026 announcements
MinLaw Guide for Generative AI in the Legal Sector
48-page non-binding guide published 6 March 2026, developed with over 20 stakeholders across law practices, in-house counsel, LegalTech providers, and academia. Three principles run through it: professional ethics (lawyers remain responsible for all work products regardless of AI involvement), confidentiality (understand how GenAI tools process and store client data), and transparency (consider disclosing AI use where it is substantial, affects cost, or involves client data handling).
Non-binding does not mean optional in practice. This guide shapes the expected standard of diligent practice.
Source: MinLaw announcement · Full guide PDF
12 initiatives across governance, infrastructure, enterprise support, and workforce development. The direction is consistent: AI adoption in legal practice is expected, financially supported, and must have a governance framework behind it.