AI & Artificial Intelligence

The federal R&D tax credit (Section 41) provides a credit of up to 20% of qualified research expenditures above a base amount, or up to 6% under the simplified method. For AI companies, qualifying activities typically include training machine learning models, developing novel algorithms, building and testing AI architectures, and improving model accuracy. Many states offer additional R&D credits. SMAART performs comprehensive studies to identify all qualifying activities and maximize your credit.

Beginning in 2022, the TCJA eliminated the ability to immediately expense domestic research and experimental (R&E) costs, instead requiring 5-year amortization for domestic costs and 15-year amortization for foreign costs. This change significantly increased taxable income for AI companies with large R&D expenditures. SMAART helps clients navigate Section 174 compliance, model the cash flow impact, and explore strategies to mitigate the effect.

Under ASC 350-40, internal-use software development costs must be capitalized once the application development stage begins, while preliminary project and post-implementation costs are expensed. For AI/ML models, determining when 'application development' begins versus preliminary research can be complex and fact-specific. SMAART helps establish proper cost classification policies that are both compliant and defensible.

Yes. We work with AI companies at every stage from pre-seed through pre-IPO. For early-stage companies, we provide fractional CFO services, board reporting, and fundraising financial support. For later-stage companies, we deliver audit preparation, Big 4 liaison services, and complex technical accounting solutions.