ENTECHNOLOGY, PRODUCTIVITY & ECONOMIC OUTLOOK · SEPTEMBER 29, 2026
Fed Governor Sees AI Productivity Potential Alongside Near-Term Cost Pressure
Federal Reserve Governor Lisa Cook said AI could become a major general-purpose technology, while its buildout may add near-term inflation pressure and its labor-market effects could arrive unevenly.
What Governor Cook said on September 28
Source publication date: September 28, 2026. Federal Reserve Governor Lisa Cook described artificial intelligence as a potentially transformative general-purpose technology. She said adoption could improve productivity and expand economic capacity over time, while the investment required for data centers, energy, equipment, worker training, reorganization, and new processes may add to near-term demand and inflation pressure. The remarks express one policymaker's analysis, not a new regulation, financing program, or guarantee of results.
- 0171% of users reported productivity gains
- 02Implementation costs can arrive first
- 03Finance measurable operating outcomes
Small firms are participating in adoption
Cook cited the Federal Reserve Small Business Credit Survey, saying nearly half of small employer firms were using AI and 71% of those users reported increased productivity. She also emphasized that adoption is still early and that smaller firms can use lower-cost tools to compete and scale. Survey responses and broad economic analysis do not establish the return on a particular software, equipment, data, or automation project.
The labor and cost effects may be uneven
Cook said current data show limited evidence of a large aggregate labor-market shift, while some occupations and entry-level roles may already face pressure. She also noted risks involving privacy, bias, fraud, cybersecurity, intellectual property, and worker transition. For a borrower, those risks can translate into implementation expense, operational disruption, control requirements, vendor dependence, and a delay between cash outlay and realized savings.
Practical borrower takeaway
Build the financing case from a defined use, vendor scope, implementation schedule, responsible owner, current process baseline, expected revenue or cost effect, training plan, data-security controls, and monthly cash forecast. Separate recurring savings from one-time costs and unproven upside, then stress-test delays, lower adoption, vendor failure, model errors, and continuing payroll. A lender will underwrite the company’s actual records, liquidity, collateral, leverage, and repayment capacity—not the economy-wide promise of AI.
Borrowers funding AI or automation should separate measurable operating gains from broad technology optimism and document implementation cost, training, cybersecurity, timing, headcount effects, and downside debt-service capacity.
Discuss a financing need →This independent summary is based on the cited source and is provided for general educational purposes only. Terms and program requirements may change.