# AI Agents Could Combat Inflation by Negotiating Better Deals Economist suggests AI may reduce costs by automating price comparisons and contract negotiations. - **Source:** [Fortune](https://fortune.com) - **Published:** 2026-09-29 - **Topic:** [Finance](https://brifing.si/c/finance) - **Original article:** - **On Brifing:** https://brifing.si/s/fa138655-25a2-449a-a744-adf21d6736c3 ## AI's dual price impact While AI capex currently drives up costs for tech resources, Wharton's Jeremy Siegel highlights its long-term deflationary potential through autonomous agents. These AI systems could negotiate better deals across banking, telecom, and insurance sectors where customer inertia maintains high prices. Current 3.4% U.S. inflation exceeds the Fed's 2% target, making cost-saving solutions politically urgent ahead of midterms, where 29% of voters prioritize economic plans. ## Muse's transactional capabilities Meta's AI agent Muse demonstrates practical applications, executing purchases like grocery orders from Instagram saves or facilitating marketplace sales. Early tests in Toronto revealed operational gaps when Muse shared a seller's address without authorization. The technology aims to automate comparison shopping and provider switching - processes consumers often neglect due to perceived effort versus savings ratio. ## Disrupting inertial monopolies Siegel compares AI's potential to the internet's price transparency impact, but with greater efficacy by removing human execution barriers. Agents could systematically move deposits to higher-yield accounts, renegotiate service contracts, and find cheaper alternatives - challenging industries that profit from consumer inaction. This represents a fundamental shift from information access to automated action in price competition. ## Consumer adoption trends OECD data shows 33% of members' citizens already use AI for financial decisions, including product selection, budgeting, and retirement planning. However, the organization warns effective use requires improved financial and AI literacy. Users must critically evaluate data requests and responses, as poor understanding increases risks alongside potential savings from automated financial management. ## Implementation challenges Early AI agent deployments reveal technical and ethical hurdles, from unauthorized data sharing to ensuring negotiation alignment with user interests. The Guardian's report of Muse's marketplace transaction highlights growing pains as systems handle sensitive operations. These issues mirror broader concerns about AI autonomy in financial decision-making processes traditionally requiring human judgment. ## Sector-specific impacts Service industries with high switching costs stand most exposed to AI-driven disintermediation. Banking and insurance particularly benefit from customer retention through inertia - a dynamic AI agents could disrupt by continuously optimizing provider relationships. This threatens incumbent revenue models built on passive consumer behavior rather than competitive service quality. ## Policy considerations As AI agents permeate consumer finance, regulators must balance innovation with protections. The OECD emphasizes literacy programs to prevent exploitation through uninformed automation. Simultaneously, policymakers might address how AI-negotiated pricing could affect market stability, especially if widespread adoption creates synchronized purchasing behaviors across consumer populations. ## Macroeconomic implications Widespread AI agent adoption could structurally lower service sector inflation by constantly pressuring providers on price. This complements traditional productivity gains from automation, adding a demand-side mechanism to supply-side efficiencies. However, the transition may prove disruptive for businesses reliant on customer inertia, requiring adaptation to more dynamic price competition. ## Sources covering this story - https://fortune.com/2026/09/29/ai-disinflationary-increased-competition-agents-siegel/ — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._