Why sales tax is especially complex for AI companies
As soon as GPT-3 was released in June 2020, the tech industry underwent a seismic shift—AI would power the next wave of innovation, impacting everything from healthcare to entertainment.
Founders and VCs moved quickly, but regulatory authorities have been slower to adapt. Most states are classifying AI services within pre-existing frameworks written for software as a service (SaaS), digital goods, data processing, and information services, creating gray areas around how AI services should be taxed.
AI billing models, often based on tokens, API calls, and compute, don’t fit neatly into legacy tax frameworks designed for physical goods or flat-rate SaaS, and bundled offerings can complicate tax treatments on a single invoice.
Sales tax complexity only increases as companies scale. When companies grow from <$1M to >$50M in revenue:
- Filing requirements increase 47x
- US jurisdiction exposure grows from 1 to 36 jurisdictions—a 3,600% increase
- Taxable revenue more than doubles, from 20% at <$1M to 47%+ at >$50M
While no state has created a dedicated “AI tax” category, certain states have introduced early AI-specific guidance. Indiana, for example, concluded that generative AI chatbot services accessed via web or API are not subject to Indiana sales tax, with Illinois passing a similar ruling. New York, on the other hand, has proposed legislation that would impose tax-like surcharges on corporations using AI to displace workers—a signal that dedicated AI tax regimes may be coming.
The sales tax landscape is continuing to evolve, and it’s critical for AI companies to develop a sales tax approach that maintains flexibility as they scale. Failing to comply with sales tax can have consequences beyond audit risk.
The cost of delaying compliance
Delaying sales tax compliance can impact key steps to growth—such as fundraising, M&A, and international expansion.
Direct financial exposure
Digital businesses that delay compliance accrue an average of 4.3% of total revenue in tax compliance liabilities and penalties. In major tech hubs like New York, exposure can climb as high as 11.3%. Back taxes typically include both uncollected tax plus interest and penalties, making it critical for high-growth AI companies to manage compliance before penalties get out of control.
Operational drain on finance teams
Teams that don’t invest in sales tax compliance solutions are left managing sales tax manually. A $10M ARR company will spend roughly 25–30 hours per month on manual tax work. With filings increasing rapidly as companies scale, manual processes quickly become unmanageable and result in errors. Before long, strategic finance work like margin analysis, pricing optimization, and growth planning goes by the wayside, replaced by tedious compliance work.
Audit risk
Audit surface increases drastically as filing volume increases. This is particularly important for AI companies, who face heightened audit risk due to the ambiguous tax treatment of novel services. If you don’t have audit-ready documentation and defensible tax determinations, you’ll be forced to reconstruct tax history with an auditor over your shoulder—a scenario no finance professional wants to be in.
Diligence and M&A delays
The AI landscape is expanding rapidly, and fundraising and M&A are a significant step in many high-growth companies’ trajectories. Sales tax exposure is one of the most common issues flagged in financial due diligence during raises and acquisitions. Any back-tax liabilities can reduce valuations, sometimes meaningfully enough to kill deals outright.
Lost expansion velocity
Recent Anrok research uncovered that tax compliance bottlenecks can delay international expansion by 3-4 months, a window that can be detrimental in AI, as technology dynamics change from quarter to quarter. Finance teams that don’t have sales tax automation in place often block or slow regional launches as manual processes catch up.
Sales tax best practices for AI companies
Sales tax laws for AI may be evolving before us, but finance and accounting teams shouldn’t wait for clear rules before taking action. Here are 3 best practices you can take as the AI sales tax landscape develops.
1. Start compliance work before you think you need to
New companies generally employ manual sales tax compliance processes. But for high-growth teams, manual workflows break down sooner than you’d think. Anrok research has found that the automation ROI breakpoint begins at $5M ARR, or around 39 filings per year, and that automation becomes essential for compliance at $10M ARR, or about 49 filings per year.
With many VC-backed AI companies under pressure to scale ARR quickly, it’s critical for finance and accounting teams to invest in compliance work early. Getting the right tools and systems in place now prevents exponential remediation costs later—back-tax cleanup is far more expensive than forward-looking compliance.
2. Monitor economic nexus continuously, not quarterly
Every state sets its own economic nexus threshold. While some states set a revenue-only threshold, like California, which uses a threshold of $500,000 in annual sales, others set a threshold in combination with a transaction trigger, like Georgia, which uses a threshold of $100,000 in annual sales or 200 transactions.
Varying nexus laws create complexity for high-growth AI companies. Finance and accounting teams need to keep track of multiple states, and a single enterprise deal or sudden viral usage can trigger nexus in different states overnight.
Investing in a real-time nexus monitoring solution identifies new obligations before penalties accrue, helping AI companies scale with confidence.
3. Classify your AI products correctly from day one
Review your product line and determine whether each offering is SaaS, digital good, IaaS, PaaS, professional services, or a bundle. Document your reasoning for each classification—rationale to reference in later audits and diligence. Keep these classifications up to date as products change and jurisdictions enact new digital laws.
Frequently asked questions
Are AI services taxable?
AI sales tax depends on the jurisdiction and how the service is classified. Many US states tax SaaS. AI outputs may fall under SaaS, digital services, data processing, or information services depending on how the product is structured. Internationally, most VAT/GST regimes tax digital services sold to local customers regardless of the provider's physical location.
How is AI taxed differently from traditional SaaS?
The core rules are similar, but AI complicates sourcing (where the compute happens vs. where the customer is), characterization (is an API call a service, a digital good, or something else), and bundling. Tax authorities are still writing guidance specific to AI, so the safest path is conservative classification with clear documentation.
When do we need to start collecting sales tax?
You should start collecting sales tax as soon as economic nexus is established in a taxing jurisdiction. For most US states, this is triggered by crossing a revenue or transaction threshold. For VAT/GST jurisdictions, it can be triggered by the first sale to a local customer. Registration and collection generally must begin within 30–60 days of threshold crossing, depending on the jurisdiction.
What happens if we've already missed filings or under-collected?
You can submit a Voluntary Disclosure Agreement (VDAs), which can reduce penalties and limit look-back periods, or register and pay taxes back directly. Acting proactively before an audit notice almost always results in better outcomes.
Do usage-based and token-based billing models change tax treatment?
Billing model type doesn't typically change whether tax applies, but it does affect how tax is calculated and reported. Systems need to handle variable transaction amounts, mid-period usage, and per-event taxation.
How does international expansion affect our obligations?
Each country has its own registration thresholds (some are $0), rates, and filing cadences. Many jurisdictions now require non-resident digital providers to register, collect, and remit as if they had physical presence. If you’re planning international operations, it’s best to partner with a sales tax automation platform that can support your expansion.
Who owns sales tax compliance internally—finance, tax, or legal?
At early stages, finance typically owns it as a part-time responsibility. By $10M+, most companies need a dedicated owner or automation platform. Cross-functional coordination with legal and RevOps is essential for product classification and nexus decisions.
How do we handle sales tax during fundraising or M&A diligence?
It’s best practice to get a current nexus analysis, filing history, and exposure estimate ready before the process starts. Acquirers and lead investors will ask for jurisdiction-by-jurisdiction compliance status. Sales tax automation platforms enable you to generate this information automatically.
Choosing the right sales tax software for your business
AI companies face a steeper compliance curve than any other business segment. They’re growing faster than companies in SaaS and eCommerce while also navigating an uncertain, constantly-evolving sales tax landscape.
Still regulators have little forgiveness. Failing to manage sales tax to the best of your ability will result in regulatory consequences—even if the rulebook is still being written.
Choosing the right sales tax software will help you automate manual work and simplify processes so that you can spend more time on strategic decisions. As you evaluate providers, focus on the criteria that matter most for AI businesses—native integrations with modern billing systems, support for AI-native billing models, global VAT/GST coverage, real-time nexus monitoring, audit support, and pricing that scales with your growth.
For a deeper dive into leading platforms available, you can check out our analysis of the 7 best sales tax automation software for AI businesses.
.png)
%20(1).webp)
