Find a TenderPlanned procurementPreliminary market engagement notice
Published 19 Nov 2026, 12:52 GMT

Official UK procurement notice

Exploration of AI & related technologies in analysis, design & implementation of tax legislation

Authority: HM Revenue & Customs (Central government)

ServicesSME suitable
Sectors: IT Services & Software (72)Research & Development (R&D) (73) Delivery: Delivery location not published
Planned pipeline valueValue not publishedpipeline notice, not a live tender
Tender notice expected Date not publishedfrom the planned procurement notice
Contract term 2 yrsends 31 Mar 2029 · starts 1 Apr 2027
Notice summaryFrom the published notice

HM Revenue & Customs (HMRC) is seeking to engage with suppliers to understand the current capabilities of Artificial Intelligence (AI) and related technologies in supporting the analysis, design and implementation of tax legislation. HMRC's strategic ambition is to deliver a tax system that is simpler for customers, easier to administer and more resilient to future change. Recent exploratory work has indicated that AI and companion technologies have potential to support tax simplification and future policy design, although further evidence is required to assess available solutions, their maturity and the feasibility of their application within HMRC.

What is being bought

Published requirement description

HM Revenue & Customs (HMRC) is seeking to engage with suppliers to understand the current capabilities of Artificial Intelligence (AI) and related technologies in supporting the analysis, design and implementation of tax legislation.

HMRC's strategic ambition is to deliver a tax system that is simpler for customers, easier to administer and more resilient to future change. Recent exploratory work has indicated that AI and companion technologies have potential to support tax simplification and future policy design, although further evidence is required to assess available solutions, their maturity and the feasibility of their application within HMRC.

We want to understand how emerging technologies could support the tax policy lifecycle, from analysing complexity in the existing system and designing simpler policy and legislation, through to expressing and implementing tax rules more consistently across operational and digital systems.

This engagement is intended to inform HMRC's understanding of available capabilities and does not represent a commitment to procure any goods or services or to implement any specific solution. However, its aim is to establish whether sufficiently mature capabilities exist to justify one or more bounded HMRC test and learn exercises.

As part of this exploratory activity, we want to understand current market capabilities, emerging technologies and supplier experience in two areas:

Workstream 1: AI and Tax Simplification

Problem statement: How can HMRC use AI to improve the way we analyse, design and simplify tax rules, and can we do so in a way that delivers sufficient benefit to justify the additional technology and resource involved?

HMRC wishes to understand whether AI-enabled tools can assist in analysing large volumes of legislation, guidance and related materials to identify potential opportunities for simplification, and whether AI can make existing complexity easier for users to understand and navigate without requiring legislative change.

Areas of interest include, but are not limited to:

  • Identifying ambiguity, duplication and unnecessary complexity within legislation;
  • Mapping legislative dependencies, cross-references and interactions;
  • Highlighting areas where legislation may be difficult to navigate, interpret or administer;
  • Supporting legislative review and simplification exercises;
  • Assisting officials in developing options for legislative reform; and
  • Generating insights that may not be readily apparent through conventional manual review.

In addition, we are also interested in proposals which could:

  • Present complex legislation and rules into simpler, accessible guidance or explanations, while accurately reflecting the underlying legislative requirements;

HMRC is interested in understanding existing products, technologies and emerging capabilities, including examples of deployment within government, regulatory or legal environments.

Workstream 2: Rules as Code

Problem statement: How can HMRC use Rules as Code (RaC) to improve the way we design, implement and maintain tax rules, and can AI reduce the effort and complexity involved in creating, translating and maintaining machine-consumable rules?

For HMRC, RaC offers the potential to support clearer rulemaking, both by applying the discipline when developing new policy and legislation and retrospectively to existing rules, while strengthening the connection between policy, law and delivery. Around 90% of digital returns received by HMRC are submitted using third-party software, across a market of approximately 3,000 software providers. Publishing machine-consumable rules could provide a common, testable and verifiable basis from which software providers, agents and businesses can implement tax rules, rather than each independently translating the underlying legislation and guidance.

Our work through the OECD's RaC project has found that translating natural-language rules into machine-consumable representations can also help expose ambiguities, unintended interactions and missing information. Therefore, applying the discipline of RaC could also help identify opportunities to clarify and simplify tax rules.

While RaC is not itself AI, creating, testing and maintaining machine-consumable rules can be resource-intensive and require significant specialist input. We want to understand whether AI can reduce this burden by supporting the translation, modelling, testing and maintenance of machine-consumable rules.

Areas of interest include:

  • The role of AI in supporting the extraction of legislative rules, concepts and decision logic;
  • Translation into structured, machine-consumable representations;
  • Identification of ambiguities, assumptions and data requirements;
  • Testing, validating, and maintaining machine-consumable rule sets;
  • Version and change management as legislation and interpretations evolve; and
  • Interoperability between tools, standards and systems.

HMRC is particularly interested in approaches that maintain legal accuracy, transparency and expert policy, legal and operational oversight, rather than replacing human judgement.

Outcomes of Interest

Our aim is to establish whether sufficiently mature capabilities exist to justify one or more bounded HMRC test-and-learn exercises and, if so, which problems, approaches and technologies offer the strongest basis for testing. The evidence gathered through this engagement will help HMRC determine where further experimentation could add most value and shape the scope and design of any subsequent test-and-learn activity.

Through this engagement HMRC wishes to understand:

  • The extent to which suppliers believe AI can support legislative simplification and RaC adoption in practice and the maturity of the market;
  • The current capability of available products and services;
  • Examples of successful deployment in legislative, regulatory or policy environments;
  • Delivery models and implementation approaches;
  • Limitations, risks and dependencies; and
  • Potential approaches for future test and learn activity.

This engagement is intended to inform HMRC's understanding of available capabilities and does not represent a commitment to procure any goods or services or to implement any specific solution.

While this is being undertaken as a preliminary market engagement exercise, HMRC is also interested in hearing from the wider expert community, including academics, researchers and other organisations with relevant expertise in AI, tax simplification and Rules as Code.

CPV: IT services: consulting, software development, Internet and support · Research and development services and related consultancy services

More open work like this

Other notices open now · live search, filters set

The lotProcurement structure

single lot · no lot value published

Pipeline intelligence · planning stage: how this market runs before the tender is published

How HM Revenue & Customs buys IT services: consulting, software, internet (CPV 72), and who usually turns up

HM Revenue & Customs’s published record next to the market for the same work. Frameworks and DPS are left out: they draw different bid counts. Whole-market view: IT services: consulting, software, internet (CPV 72) →

1

How the contest usually goes

Winner shares use published awards; bid and price charts use contested lots.

Question for the bid
This buyerHM Revenue & Customs · all categories
The marketIT services: consulting, software, internet (CPV 72) · Greater London (authority’s region)
What to do with it
How often do SMEs win?
i

The counts under this question show how many awards have a known supplier size.

Buyer: 14 of 52 known-size awards · market: division 72 · Greater London (authority’s region), n=943
27%49%

Small firms have won here. Neither number says whether you qualify; it describes who has been winning.

How often do local suppliers win?
i

Local means the winner is registered in the region where that award was delivered. The market figure counts awards delivered in Greater London (authority’s region).

Buyer: 4 of 4 known-location awards · market: division 72 · Greater London (authority’s region), n=712
4 of 439%

Most of this buyer’s recorded winners were registered in the region where that award was delivered.

How long to a decision?
i
Buyer: 11 decided notices · market: division 72 · Greater London (authority’s region), n=74
87 days66 days

Half of these awards have an award date within 87 days of the bid deadline.

How often single-bid or direct?
i

Both shares are minimums: some awards have no bid count.

Buyer: at least 38 of 54 procurements with awards had no recorded rival · market: division 72 · Greater London (authority’s region), n=1,064
70%65%

Direct awards or one recorded bid are common here. Check the route before spending time on a bid.

How many bids did past lots attract?

i
This buyer18 lots with bid counts · all categories
17% had one bid
The market107 lots with bid counts · division 72 · Greater London (authority’s region)
7% had one bid
% of lots with a published bid count: this buyer (18) / the market (107) · select a colour for counts
1 bid17% / 7%
Buyer: 3 of 18 lots with published bid counts. Market: 8 of 107 lots with published bid counts.
2–356% / 32%
Buyer: 10 of 18 lots with published bid counts. Market: 34 of 107 lots with published bid counts.
4–622% / 31%
Buyer: 4 of 18 lots with published bid counts. Market: 33 of 107 lots with published bid counts.
7+6% / 30%
Buyer: 1 of 18 lots with published bid counts. Market: 32 of 107 lots with published bid counts.

How did awards compare with estimates?

i
This buyer11 lots with estimate and award · all categories
27% below estimate
The market79 lots with estimate and award · division 72 · Greater London (authority’s region)
41% below estimate
% of lots with an estimate and award: this buyer (11) / the market (79) · select a colour for counts
Below estimate27% / 41%
Buyer: 3 of 11 lots with an estimate and award. Market: 32 of 79 lots with an estimate and award.
Within ±2%64% / 39%
Buyer: 7 of 11 lots with an estimate and award. Market: 31 of 79 lots with an estimate and award.
Above estimate9% / 20%
Buyer: 1 of 11 lots with an estimate and award. Market: 16 of 79 lots with an estimate and award.
2

Assess each lot
i

Open a lot for its forecast, this buyer’s previous supplier and possible competitors, from published award history, not a bid list

Lot #1CPV 7224 monthsSME suitableEstimate not published

This lot lists more than one class in CPV 72. The comparison uses that group.

Expected competition
i

Each lot shows its own forecast based on its own category; forecasts for different lots of the same notice are not comparative. The bar shows 107 past lots.

—No model forecast yet7% chance of a single bid
Past bids on 107 similar lots
1 bid7%
8 of 107 similar past lots received 1 bid (7%).
2–332%
34 of 107 similar past lots received 2–3 (32%).
4–631%
33 of 107 similar past lots received 4–6 (31%).
7+30%
32 of 107 similar past lots received 7+ (30%).
Comparable past lots usually drew 4 bids.
Commercial price forecast
i

No comparable awards for this lot yet.

—Estimate not published
The buyer published no estimate for this lot, so there is nothing to forecast against. A price prediction here would be a number about a number we do not have.

Who has supplied this buyer on similar work?
i

No earlier contract of this buyer matches this lot on the published record.

Who could compete for this lot?
i

Candidate suppliers for this notice have not been worked out yet, so this is not a verdict that nobody fits.

About this data
  • Estimates, labelled. The price range, possible competitors and earlier-work matches are our estimates from past awards, not published facts, win chances or price advice. The buyer’s own record and the wider market are shown apart, never mixed.
  • Source. Official UK public procurement notices (Find a Tender and Contracts Finder), reused under the Open Government Licence v3.0, updated through the day and rebuilt nightly. Not every figure uses both feeds, and each one shows how many records it rests on: see the methodology. Spotted something wrong? Tell us.

Procedure terms & legal framework

From the official record
Legal basis
UKPGA · 2023/54 Procurement Act 2023
Procurement category
Services
Notice status
Planned procurement
Procurement threshold
Unspecified in notice
Regulatory regime
Standard
Commercial structure
Stand-alone contract (no framework)
Competition type
Competitive procedure
GPA / WTO covered
Unspecified in notice
Recurring procurement
Unspecified in notice
First published
19 Nov 2026, 12:52 GMT
Last source update
19 Nov 2026, 12:52 GMT
FTS notice number
2026/S 000-109724
Latest notice
Preliminary market engagement notice (UK2)
Classification (CPV)
IT services: consulting, software development, Internet and supportResearch and development services and related consultancy services
Delivery area
Not published
Official registry OCID
ocds-h6vhtk-07bf7c Find a Tender

How it was run

Official notice
Submission deadline
—

Timeline

19 Nov 2026 → 1 Apr 2027 · 3 dates
  1. Planning notice published11h ago
    19 Nov 2026, 12:52 GMT
  2. Today
    20 Nov 2026
  3. Market engagementin 19 days
    9 Dec 2026, 23:59 GMT

    HMRC will hold a market engagement event with interested suppliers to discuss the challenges and opportunities outlined above. This event will be hosted digitally on 9th December 2026 at 12pm GMT. Please contact [email protected] to express your interest in attending.

    In the event, we will:

    • Provide further detail on HMRC's objectives with respect to;
    • Using AI to aid tax simplification; and
    • Rules as Code.
    • Set out our emerging thinking on potential focus areas for future test and learn activities.

    We would also like to:

    • Explore current and emerging market capabilities;
    • Seek views on technical, legal and operational constraints;
    • If appropriate, invite suppliers to briefly outline how they would approach the challenges identified by HMRC, where they are comfortable in doing so; and
    • Provide an open forum for suppliers to ask questions and seek clarification on HMRC’s objectives and the challenges we are seeking to address.

    Following the event, participants are invited to provide written responses describing their solutions, relevant experience, and views on market maturity. We may then invite contributors to discuss their views in more detail on a one-to-one basis.

    We are particularly interested in hearing from small and medium-sized enterprises, UK-based suppliers and organisations with experience in legal technology, regulatory technology (RegTech), legislative analysis and Rules as Code.

    The information gathered from this exercise will be used to inform our next steps with respect to a potential test and learn exercise.

    Participation in this engagement exercise will not confer any advantage in any future procurement process. Any future procurement activity will be subject to separate approval and procurement procedures.

    Key Questions for Suppliers

    Tax Simplification

    1. What tools, capabilities or methods can you provide to support the analysis of legislation, regulation and policy documentation?
    2. How would your solution assist in identifying complexity, or perceived complexity, and providing options for simplification?
    3. What evidence exists that your solution can support tax simplification or regulatory reform?
    4. What role does AI play in your approach? Please distinguish between capabilities available today, and those that are experimental or under development, and explain where human review remains necessary.
    5. What are the technical, legal, data, organisational and operational dependencies or limitations of your approach?
    6. If HMRC wanted to test this capability through a small, bounded proof of concept, what type of rule would you recommend testing, what would the outputs be, and what evidence would demonstrate whether the approach had added value?

    Rules as Code

    1. What tools, capabilities or methods can you provide to support the translation, modelling, testing, use and maintenance of UK legislation or policy rules as machine-consumable representations?

    a. In particular, how would your approach accommodate characteristics of UK tax legislation, and how would different approaches to translation enable different uses of the resulting machine-consumable rules?

    1. How does your solution support the:

    a. translation of legislation into machine-consumable rules?

    b. translation of non-legislative materials (e.g. case law, guidance, force of law notice) into machine-consumable rules?

    c. maintenance of translated rules following change?

    1. How does your solution distinguish between requirements that are explicit in legislation, guidance, and/or case law and those that need to be inferred or interpreted, and how are these surfaced for human review rather than automatically resolved?
    2. What role does AI play in your approach? Please distinguish between capabilities available today, and those that are experimental or under development, and explain where human review remains necessary.
    3. What are the technical, legal, data, organisational and operational dependencies or limitations of your approach?
    4. If HMRC wanted to test this capability through a small, bounded proof of concept, what type of rule would you recommend testing, what would the outputs be, and what evidence would demonstrate whether the approach had added value?
  4. Contract startsin 131 days
    1 Apr 2027, 00:00 BST

Notice history

1 event · 1 release on Find a Tender
  1. Preliminary market engagement notice published
    19 Nov 2026, 12:52 GMT
    UK2

Commercial outcome & contract awards

Pipeline notice
Awards (0)
No award published
Contracts (0)
No contract published
Bid statistics
No aggregate bid statistics published

Contracting authority & participating public bodies

HM Revenue & Customs

Contracting authorityPublic authority - central government GB-PPON: PVMW-8599-JZNJ 100 Parliament Street, London, London, SW1A 2BQContact name, email, phone

Official documents & specifications (1)

Official links; attachments are not copied

Notices published for this procedure