Future Scope of AI in Corporate Taxation and Law (2026)

November 5, 2026No Comments
future-scope

Introduction

Artificial intelligence is moving from experiment to everyday tool in tax departments, law firms and tax authorities. In Pakistan, the Federal Board of Revenue (FBR) has publicly discussed using AI to cross-check returns, and the federal cabinet approved a National AI Policy in July 2025. Professionals who understand both the technology and the rules will be in a strong position.

This matters whether you are an accountant, tax consultant, lawyer, company secretary, finance manager or student. Routine work such as data entry, document sorting and first-draft research is increasingly automated. Judgement, interpretation and accountability remain human responsibilities. If you want a shorter primer first, see our guide to AI in taxation: tools, risks and opportunities.

This article explains what AI means in corporate taxation and law, how it is already being used, what Pakistan’s tax authority is doing, where the risks lie, and which skills and careers will matter next. It is written for practitioners and learners, not as legal or tax advice.

Quick Answer: What Is the Future Scope of AI in Corporate Taxation and Law?

AI will increasingly handle data-heavy tasks in corporate taxation and law, such as return checks, audit risk scoring, contract review and legal research. Professionals will still be needed to verify outputs, interpret law, advise clients and take responsibility. The strongest future belongs to those who combine technical knowledge with AI literacy.

Key points:

  • AI is strongest at pattern detection, document analysis, drafting and summarising.
  • Tax authorities are adopting AI for compliance monitoring and audit selection.
  • Accuracy, confidentiality and explainability are the main risks.
  • Roles are shifting towards review, advisory, data and governance work.
  • Practical training in both tax or law and AI tools is the best preparation.

What Is AI in Corporate Taxation and Law?

AI in corporate taxation and law means using software that can learn from data, read language and generate content to support tax and legal work. Examples include tools that read invoices, flag unusual transactions, summarise judgments, compare contract clauses or draft a first version of a legal memo.

How Is AI Different From Tax Automation and RPA?

Traditional automation follows fixed rules. If a field is blank, the software stops; if a tax code matches, it applies a set rate. Robotic process automation (RPA) copies repetitive clicks and keystrokes between systems. AI goes further because it can interpret messy inputs such as scanned documents, emails or long contracts, and it can produce probabilistic outputs rather than guaranteed answers.

That difference matters. A rule-based system is predictable and easy to audit. An AI system can be more flexible but needs testing, supervision and clear records of how it was used. Our comparison of manual and AI-assisted tax filing shows where each approach fits.

Which Technologies Power AI in Tax and Law?

  • Machine learning: finds patterns and anomalies in transactions, such as unusual withholding or sales tax entries.
  • Natural language processing: reads statutes, notices, judgments and contracts.
  • Large language models (LLMs) and generative AI: draft text, answer questions and summarise documents.
  • Intelligent document processing and OCR: extract data from invoices, bank statements and certificates.
  • Predictive analytics: estimates tax exposure, cash tax and audit risk.
  • Explainable AI: helps users understand why a model produced a result, which is vital for regulated work.

Why Is AI Becoming Important in Tax and Legal Work Now?

Three pressures are pushing adoption. First, data volumes have grown because of e-invoicing, digital payments and online filing. Second, regulators want better compliance with fewer manual steps. Third, clients expect faster and more affordable advice.

Globally, tax authorities are already ahead of many businesses. The OECD’s report on tax administration digitalisation says more than 70% of the tax administrations it covers use AI to improve efficiency, for example in compliance management and taxpayer services. The report draws on the OECD Forum on Tax Administration’s inventory of 54 members.

In Pakistan, the government set out a national direction when the cabinet approved the National AI Policy 2025 on 30 July 2025. Media reports at the time said it targets training one million AI professionals by 2030 and creates an AI Council to oversee implementation.

How Is AI Changing Corporate Taxation?

How Does AI Change Tax Compliance and Return Preparation?

AI can extract figures from invoices and ledgers, match them against bank and withholding records, and highlight mismatches before a return is filed. This reduces keying errors and shortens review time. It does not remove the need for a qualified person to confirm the treatment of each item.

Businesses preparing for FBR’s digital environment should also understand FBR IRIS 2.0 features and the FBR digital invoicing system, because clean, structured data is what makes AI useful. Our guides to e-invoicing in Pakistan and digital invoicing and tax compliance explain the compliance side.

How Can AI Improve Corporate Tax Planning and Forecasting?

Planning tools can model scenarios, such as how a restructuring, new contract or change in the Finance Act might affect tax cost. Machine learning can improve cash tax forecasting by learning from past patterns. The output is a starting point for discussion, not a recommendation. A tax adviser still has to test assumptions against current law. For the Pakistani context, read corporate tax in Pakistan, corporate tax planning and the latest budget tax changes.

How Is AI Used in Tax Audits and Risk Management?

This is where change is fastest. Authorities can combine return data with third-party information to identify returns that look inconsistent. Businesses can use similar tools internally to test their own data before an audit. Our explainer on digital tax audits and AI covers the mechanics, and our tax audit guide for Pakistan and FBR audit notice guide explain how to respond if you are selected. If a dispute follows, the route is explained in tax appeals from Commissioner to ATIR.

Where Does AI Help With Withholding Tax and Reconciliation?

Withholding tax generates large volumes of transactions and statements. AI-assisted reconciliation can match deductions with payments and certificates and flag gaps before a statement is filed. Read more about withholding tax on bank transactions and the Section 165 withholding statement.

How Does AI Help With Sales Tax and Indirect Tax Compliance?

Sales tax compliance depends on accurate invoices, correct classification and reconciliation of input and output tax. AI can help read invoices, match them with ledgers, flag missing or inconsistent entries and prepare data for return filing. As e-invoicing expands, structured invoice data makes this more reliable.

The law and filing steps still need human confirmation. To understand the compliance basics, read our guides on registering for sales tax in Pakistan, filing sales tax returns and the benefits of digital invoicing for FBR compliance. Professionals who want deeper training can review the Master Sales Tax course.

What About Transfer Pricing and International Tax?

Transfer pricing and cross-border tax involve comparables, documentation and treaty analysis. AI can help search and organise this material, and summarise treaty provisions. The professional still decides whether a comparable is genuinely comparable and whether a treaty position is defensible. Businesses with foreign operations should also review cross-border tax compliance.

What Is Pakistan’s Tax Authority Doing With AI?

According to Arab News, the FBR is preparing to expand AI checks on tax returns to identify undeclared income and assets. The report says models will compare taxpayers’ declarations with property, banking and other financial records, as part of an IMF-backed reform programme. It also reports a federal tax collection target of Rs15.26 trillion for the financial year ending June 2027.

Business Recorder reported that the FBR Chairman presented the Pakistan New Tax Operating Model to the National Assembly Standing Committee on Finance. As described in that report, a National Faceless Audit Wing would conduct risk-based audits using a Central Data Hub, and AI-generated discrepancy reports would feed a staged “green channel” process in which taxpayers can pay based on the report. The committee chairman asked the FBR to explain the trigger points the AI system would use to start audits. Reports have indicated a phased rollout, so check the FBR website for the current status before relying on any timeline.

What this means for businesses:

  • Keep records consistent across returns, bank statements, withholding statements and financial statements.
  • Treat discrepancy notices seriously and respond with documents, not assumptions.
  • Expect more questions about how automated selection works and what safeguards exist.
  • Make sure someone in the business understands the digital filing environment. Our overview of FBR’s digital transformation is a good starting point.

What Should a Business Do If an AI-Generated Discrepancy or Notice Arrives?

A discrepancy report or notice is a request for explanation, not a final decision. The right response is calm, documented and timely.

  1. Read the notice fully and note the tax year, the legal provision cited and the response deadline. Our FBR notices explained guide helps you identify the type of notice.
  2. Compare the alleged discrepancy with your returns, bank records, withholding statements and financial statements.
  3. Identify whether the issue is a genuine error, a timing difference or a data mismatch on the authority’s side.
  4. Gather supporting documents such as invoices, contracts, certificates and reconciliations.
  5. Prepare a clear written reply. For withholding-related notices, see how to reply to an FBR Section 161 notice. For other notice types, read the Section 114 notice guide.
  6. If a return needs correction, learn the process in how to file a revised return on IRIS 2.0 and review common IRIS 2.0 filing errors.
  7. Keep copies of everything you submit, and consult a qualified tax professional for complex or high-value matters.

Prevention is cheaper than correction. Review common tax mistakes made by Pakistani businesses and our overview of Income Tax Ordinance compliance.

How Is AI Changing Corporate Law and Legal Services?

How Does AI Support Contract Management and Drafting?

AI can extract key dates, obligations and clauses from large contract sets, compare them with a preferred template and suggest redlines. This is especially useful for repeat agreements. A lawyer must still decide whether a clause fits the commercial deal and applicable law. For the legal background, see corporate law in Pakistan and how corporate law and taxation work together.

How Does AI Help With Legal Research and Due Diligence?

Research tools can summarise judgments, find related authorities and highlight inconsistent holdings. In due diligence, AI can sort data-room documents and surface unusual terms, such as change-of-control clauses or unusual tax indemnities. Results always need checking against the original source.

How Can AI Support Corporate Governance and Regulatory Compliance?

Compliance teams can use AI to monitor regulatory updates, track filing calendars and prepare board papers. For company law obligations, our guides on company registration and compliance in Pakistan and SECP annual return filing show the underlying duties. The SECP has also said it is working on an AI-based system to speed up company registration, according to a July 2026 news report, so watch its official announcements.

What Is the Difference Between AI-Assisted and Traditional Tax and Legal Work?

TaskTraditional ApproachAI-Assisted ApproachHuman Role
Return ReviewManual checking of ledgersAutomated mismatch detectionConfirm treatment and sign off
Legal ResearchKeyword search and readingSemantic search and summariesVerify sources and apply law
Contract ReviewClause-by-clause readingClause extraction and flaggingJudge commercial and legal fit
Audit ResponseManual document gatheringFaster data retrieval and draftsStrategy and representation
PlanningSpreadsheet modelsScenario modellingInterpret law and advise client

For tax advice specifically, our human vs AI tax advisory comparison and AI vs tax consultant comparison go deeper.

Which AI Tools Do Tax and Legal Professionals Use?

No single tool covers every tax or legal task. In practice, professionals combine several categories of software. Before adopting any of them, check what data the tool stores, which jurisdictions it covers, and whether it can cite its sources. Many tools are built mainly around US or UK law and may not reflect current Pakistani rules, so verify outputs against official sources.

What Can General-Purpose AI Assistants Do?

Assistants such as ChatGPT, Microsoft Copilot, Google Gemini, Claude and Perplexity can draft, summarise, explain and organise information. They are useful for first drafts of client emails, plain-language explanations and research outlines. They are not substitutes for the Income Tax Ordinance, the Sales Tax Act or the Companies Act, so every section number, rate and deadline must be checked at source. Our guides to ChatGPT for tax professionals and AI tools for tax consultants in Pakistan show how to use them carefully.

Which Tools Support Legal Research and Contract Review?

Legal AI assistants, such as Thomson Reuters CoCounsel, and contract lifecycle platforms are designed for research, clause comparison and contract tracking. They work best when the underlying legal database matches your jurisdiction. Law students and graduates who want to understand the subject behind these tools can start with corporate law courses in Islamabad.

future-scope
future-scope

How Do Document Processing and Automation Tools Help?

Intelligent document processing and OCR tools, such as ABBYY, read invoices, bank statements and certificates. Robotic process automation platforms, such as UiPath, move that data between systems. Together they reduce manual entry, but they depend on clean source documents. If you want to build skills in this area, see the AI data analyst course guide for Pakistan and advanced Excel training.

Where Do Accounting and ERP Platforms Fit?

Accounting and ERP systems such as Xero, QuickBooks, Odoo, SAP, Oracle and Microsoft Dynamics 365 hold the transaction data that tax and legal teams rely on. AI features inside these platforms can assist with categorisation, reconciliation and reporting. Learners can explore the complete Xero accounting guide and the Xero training course. For international filing work, see the USA tax filing software professionals use.

ChatGPT or Microsoft Copilot: How Should a Tax Professional Choose?

Instead of asking which is “best”, compare them against your own requirements:

  • Data handling: Does the plan allow you to restrict how client data is stored or used?
  • Integration: Do you already work in Microsoft 365, or in another environment?
  • Source transparency: Can the tool show where an answer came from?
  • Administration: Can your firm manage users, access and audit trails?
  • Cost model: Is pricing per user, per usage or bundled with other software?
  • Training: Will staff need structured guidance to use it responsibly?

Run the same set of anonymised test questions through each option and compare accuracy before committing.

Will AI Replace Tax Accountants and Corporate Lawyers?

Unlikely in full, but the work will change. Tasks that are repetitive, rule-based and document-heavy are the most exposed. Tasks that need judgement, negotiation, client trust, ethical responsibility and representation before authorities or courts remain human-led.

In practice, professionals who use AI well may outperform those who do not. That is why a growing number of firms are asking what skills they need rather than whether the profession will disappear. Read our views in will AI replace tax consultants, why businesses still need a tax adviser in the AI era and the impact of AI on tax consultants.

What Are the Risks and Limitations of AI in Taxation and Law?

Can AI Make Mistakes in Tax and Legal Research?

Yes. Generative AI can produce confident but false answers, known as hallucinations. In the legal world, courts have sanctioned lawyers who filed AI-generated citations that did not exist. For example, the ABA Journal reported that a US federal judge sanctioned lawyers over fake case citations, with the motion citing eight nonexistent cases and a total sanction of $5,000. The lesson applies in tax too: never rely on an AI answer for a section number, rate or deadline without checking the official source. Our guide to AI tax calculation and legal risk discusses this further.

What Are the Data Privacy and Confidentiality Risks?

Tax and legal files contain sensitive client data. Entering them into public AI tools can create confidentiality and security risks. Firms should use approved tools, understand where data is stored, and set clear internal rules on what may be uploaded.

Why Do Explainability and Accountability Matter?

If an automated system flags a taxpayer or a contract clause, the professional or authority must be able to explain why. The OECD has highlighted privacy, fairness and accountability as key challenges for AI in tax administration. Responsibility cannot be delegated to software, so records of prompts, sources and reviewer sign-off are good practice.

Can AI Reflect Bias or Poor Data Quality?

Models learn from data. Incomplete or inconsistent data can lead to wrong risk scores or missed issues. Cleaning and structuring data before using AI is often more valuable than buying another tool.

What Rules and Policies Govern AI Use in Pakistan?

Pakistan’s main national direction is currently the National AI Policy 2025 rather than a single dedicated AI statute, as far as we could verify. The policy also refers to data protection, cybersecurity and alignment with international AI norms. Professionals should therefore rely on existing laws on tax confidentiality, company law, professional conduct and data security, and check official sources such as the FBR, SECP and Pakistan Code for updates. Rules in this area can change quickly, so verify before acting.

What Skills Will Tax and Legal Professionals Need in the AI Era?

  • Strong fundamentals: income tax, sales tax, withholding, company law and contracts.
  • AI literacy: understanding what tools can and cannot do, and how to check outputs.
  • Prompting and verification: asking precise questions and testing answers against primary sources.
  • Data skills: spreadsheets, dashboards and basic analytics. See AI vs traditional data analytics.
  • Accounting systems: familiarity with ERP and cloud tools such as Odoo accounting and QuickBooks.
  • Ethics and confidentiality: knowing when not to use AI.

See also the tax professional skills in demand and the software tax professionals use.

What Career Opportunities Will AI Create in Taxation and Law?

Roles are likely to grow around AI-assisted advisory, tax technology, compliance analytics, legal operations, contract management and governance. Examples include tax data analyst, compliance automation specialist, legal operations coordinator and AI-literate corporate adviser. We have not quoted salary or demand figures because reliable, current data should come from official or professional sources.

For career planning, explore new tax careers and digital compliance roles, how tax careers are evolving with technology, corporate law careers, and AI tools freelance tax consultants will use.

How Can You Learn AI Skills for Taxation and Corporate Law in Pakistan?

The most effective route is to learn the subject first and the tools second. AI is much more useful in the hands of someone who can recognise a wrong answer.

Is AI a Good Career Direction for Law and Commerce Students?

It can be, especially in roles that combine legal, tax or accounting knowledge with technology. Students should build a strong base in company law, contracts and taxation, then add AI literacy and data skills. Useful starting points include corporate law myths students should ignore, why CA students must learn practical FBR filing and whether taxation is a good career in Pakistan.

Can Working Professionals Learn These Skills Part-Time?

Many professionals upskill while employed. Plan a small weekly learning routine and apply each new skill to a real task at work. Read can I study tax while working full time and compare online vs physical tax courses to decide which format suits you.

Which Course Path Fits Your Starting Point?

What If You Live Outside Islamabad?

Learners in Karachi, Lahore, Rawalpindi, Peshawar, Faisalabad, Multan, Sialkot and other cities should compare delivery formats, schedules and certification details before enrolling. ICT has published guidance mainly for Islamabad learners, so check the ICT courses page or contact ICT to confirm current formats and admission details.

Can Business Advisers and Company Secretaries Use AI?

Yes, for research, drafting board papers, tracking compliance calendars and summarising regulatory updates, always with human review. To understand these career paths, see corporate advisory career scope in Pakistan, how corporate law boosts business advisory, the Certified Corporate Advisor course guide and the Certified Business Advisor course.

Do AI Skills Help in International Tax Careers?

They can. International clients value professionals who can combine foreign tax knowledge with efficient digital workflows. Explore USA taxation, UAE taxation, and the broader context in global tax trends and their impact on Pakistan. For remote work possibilities, read about remote tax jobs.

How Can a Firm or Tax Department Start Using AI Safely?

  1. Define one narrow use case, such as invoice data extraction or contract clause search.
  2. Check data quality and confidentiality rules before choosing a tool.
  3. Test the tool on past, anonymised examples and compare results with known answers.
  4. Require human review and keep a record of sources and approvals.
  5. Train staff on strengths, limits and ethical use. Start with ChatGPT for tax professionals and ChatGPT prompts for tax consultants.
  6. Review results regularly and expand only when accuracy and controls are proven.

Tool selection depends on your size and needs. Compare options in AI accounting software, best AI tools for accountants, AI audit tools and AI for bookkeeping. Pricing varies by vendor, so request current quotes rather than relying on estimates.

AI Tax and Legal Compliance Checklist

Use this list before relying on an AI tool for client or company work:

  • The task is clearly defined and suitable for AI support.
  • Confidential data is handled under approved tools and written rules.
  • Every law, rate, section number and deadline is checked against an official source.
  • A qualified reviewer approves the final output.
  • Prompts, sources and approvals are recorded for audit purposes.
  • Returns, statements and certificates are reconciled with source records.
  • Staff know what not to upload or automate.
  • The tool is tested on past, anonymised examples before live use.
  • Filing dates are tracked in a reliable calendar, such as the Pakistan tax calendar.
  • Policies are reviewed regularly as laws and tools change.

How Much Does AI Tax and Legal Software Cost?

Prices vary by vendor, plan, number of users and features, and they change often, so this article does not quote figures. Ask vendors for a written, current quotation and compare it with the time you expect to save.

Costs to consider:

  • Licences: per-user, per-usage or bundled subscriptions.
  • Implementation: setup, integration with accounting or document systems and data cleaning.
  • Training: staff time to learn the tool and your internal rules.
  • Verification: the time reviewers spend checking outputs, which remains a real cost.
  • Security and compliance review: assessing data protection and confidentiality risks.

Are There Free AI Tools for Tax Professionals?

Free versions of general AI assistants exist, but they may handle data differently from business plans. Do not upload confidential client files to a tool unless you understand its data terms. For learning and non-confidential drafting, free tools can be a low-risk way to practise.

Is AI Affordable for Small Businesses and Startups?

Often the best approach is to start small, with one task such as invoice data extraction or contract summarising, and measure the benefit. Training is also an investment, so see taxation course fees in Pakistan and our analysis of return on investment in tax training when planning a budget.

What Should Professionals Watch Between Now and 2030?

These are directions to monitor, not guaranteed outcomes:

  • Wider use of AI and faceless processes in tax administration, and the safeguards that accompany them.
  • More structured, machine-readable tax data from e-invoicing and digital reporting.
  • AI agents that carry out multi-step compliance tasks under human supervision.
  • Stronger professional guidance on AI use, verification and disclosure.
  • Growing demand for people who can bridge tax, law and technology.

Our outlook pieces on the future of tax advisory in Pakistan and the future of accounting after AI cover related ground.

Glossary of Key AI, Tax and Legal Technology Terms

  • Artificial intelligence (AI): software that performs tasks normally requiring human intelligence, such as reading, predicting or drafting.
  • Machine learning: a method where software learns patterns from data instead of following only fixed rules.
  • Large language model (LLM): an AI model trained on large amounts of text that can generate and summarise language.
  • Generative AI: AI that creates new content, such as text, tables or summaries.
  • Hallucination: a confident but false or invented AI output.
  • Natural language processing (NLP): technology that helps computers understand human language.
  • Robotic process automation (RPA): software that repeats rule-based tasks across systems.
  • Optical character recognition (OCR): technology that converts scanned text into editable data.
  • Intelligent document processing: tools that extract and classify information from documents.
  • Predictive analytics: using data to estimate future outcomes, such as tax exposure or audit risk.
  • Explainable AI: methods that help people understand why an AI system produced a result.
  • Faceless assessment: a process where taxpayer and officer interaction is limited or anonymised through a central system. Learn how it fits Pakistan’s wider reform in our overview of Pakistan’s tax system in 2026.

Why Choose ICT for AI-Ready Taxation and Corporate Law Training

The Institute of Corporate and Taxation (ICT) teaches taxation and corporate subjects alongside technology-focused programmes. Its courses page lists options such as the Certified Tax Advisor course, Advance Taxation and Litigation, the Company Secretary course, and tech-oriented programmes including the Certified AI Automation Agents Specialist and AI-Driven CFO Masterclass.

Learners who want to build data skills can also review the Certified Data Analyst and Odoo for Business ERP Automation programmes. You can learn more about ICT and use its certificate verification page to confirm credentials.

Frequently Asked Questions

What is the future scope of AI in corporate taxation?
AI will take on more compliance checking, data matching, risk analysis and research support, while professionals focus on interpretation, advice and accountability.

Will AI replace corporate lawyers?
Not entirely. AI can speed up drafting, research and review, but lawyers are still needed for judgement, negotiation, ethics and representation.

Can AI prepare corporate tax returns?
AI can extract data and suggest entries, but a qualified person must review the treatment and take responsibility before filing.

Is AI used by the FBR?
Media reports say the FBR is expanding AI checks on returns and has presented an AI-based operating model to parliament. Check the FBR website for current implementation details.

What are the main risks of AI in tax and legal work?
Hallucinated answers, data confidentiality, poor explainability, bias from weak data, and unclear accountability.

Which skills should tax professionals learn for the AI era?
Tax and law fundamentals, AI literacy, verification skills, data analysis, accounting software and professional ethics.

Is AI a good career direction for tax and law graduates?
It can be, particularly in roles that combine tax or legal knowledge with technology. Build core expertise first, then add AI skills.

How should a small firm start with AI?
Pick one narrow task, use approved tools, protect client data, test on past examples and always keep human review. See our AI for tax consultants in Pakistan guide.

Which AI tools should tax professionals learn first?
Start with a general AI assistant for drafting and summarising, then learn spreadsheets and your accounting platform. See best AI tools for accountants.

Is AI tax software more accurate than a human accountant?
Not automatically. AI can be faster at pattern matching, but accuracy depends on data quality, the tool and review. Human oversight remains essential.

How much does AI tax software cost?
It varies by vendor, users and features. Ask vendors for current written quotations rather than relying on estimates.

Can AI help with FBR notices?
It can help organise documents and draft a reply outline, but a qualified professional should confirm the legal position. See our FBR notices guide.

Do corporate lawyers need to learn AI?
Yes, at least enough to use it responsibly, verify outputs and advise clients on AI-related risks.

Can AI help with sales tax compliance?
It can assist with invoice reading, reconciliation and anomaly detection, but filings need professional review. Start with how to register for sales tax in Pakistan.

Conclusion

AI is already reshaping how tax authorities and professional firms work, and the shift will continue. It can cut routine effort, improve checking and widen access to information, but it cannot replace professional judgement or responsibility. The risks of errors, confidentiality breaches and weak governance are real and manageable with the right controls.

The best response is to build strong foundations in taxation and corporate law, then learn how to use AI tools carefully. To start that journey, explore ICT’s Certified Tax Advisor course for tax foundations, or the Certified AI Automation Agents Specialist programme if you want to focus on automation. You can also to ask which course suits your background.

Comments (0)

No comments yet. Start the conversation!


Leave a Reply

Your email address will not be published. Required fields are marked *

Subscribe to our newsletter for the latest updates and insights.

Stay ahead with the latest updates, insights, and events from ICT.

© 2026 ICT. All rights reserved.

Chat with us