As shifting trade policies create new friction in global supply chains, the traditional monthly reporting cycle is no longer a sufficient shield for profitability. MindBridge CFO Matthias Steinberg explores how AI is moving finance toward continuous monitoring to catch the "hidden" margin erosion that legacy ERPs consistently miss.
For the modern finance leader, the spreadsheet is no longer a shield it’s a blind spot. As geopolitical volatility triggers sudden shifts in tariff policies, the traditional monthly reporting cycle is proving too slow to catch the “death by a thousand cuts” that erodes global margins. For CFOs, the challenge is not just the cost of goods but the speed at which those costs change.
We sat down with Matthias Steinberg, CFO at MindBridge, to discuss how AI is shifting the finance function from a reactive “autopsy” of past performance to a continuous, proactive defense of profitability. Steinberg, whose background spans senior roles at IONOS and Summit Partners, argues that in an era of shifting trade policy, the “materiality threshold” is a legacy concept that may be hiding significant risk.
In a volatile trade environment, margin erosion rarely arrives as a single, headline-grabbing event. Instead, it manifests in fragments. Amended duty rates appear on invoices, suppliers introduce surcharges that sit outside the primary tariff line, and inventory journals are posted to correct earlier cost assumptions. Individually, these items are rarely material enough to trigger an alert; collectively, they can devastate a quarterly result.
Steinberg explains that traditional monthly or quarterly cycles aggregate this activity only after the period closes. By that stage, the underlying pattern has already moved through inventory and into the cost of sales. To combat this, finance leaders must shift their focus toward behavioral signals rather than headline figures. Continuous monitoring allows teams to detect inconsistencies between rate tables and actual invoices the moment they enter the ledger, identifying shifts in landed cost components before they compound.
Standard ERP systems are built on the assumption of static rules and periodic updates. However, tariff changes often move faster than a global enterprise can update its internal rate tables. This creates a “timing gap” where invoices arrive reflecting revised duties before the system logic has been updated to match.
Steinberg identifies a specific “breaking point” for legacy tools: the moment cost behavior changes faster than models and control settings can be updated. Spreadsheets and legacy forecasting models struggle when cost shifts are broad-based and uneven. A sudden 20% increase in import duties across multiple jurisdictions does not affect a single line item; it filters through thousands of transactions and product codes. Traditional models flatten these distributed changes into summary variances, often masking the true exposure.
Furthermore, the threat in the current trade environment is less about isolated clerical errors and more about structural blind spots. When impact is spread across thousands of transactions, it doesn’t trigger exception reporting. It sits quietly within working capital and inventory movements. Full-population analysis reviewing every single transaction rather than a sample is the only way to surface these hidden costs buried in complex supplier contracts.
This shift to AI-driven visibility fundamentally changes the relationship between Finance and the Chief Procurement Officer (CPO). Instead of discussing variances weeks after they occurred, finance can provide real-time data on which suppliers are absorbing costs and which are passing them on through opaque line items.
This intelligence allows for a shift from retrospective explanation to forward-looking negotiation. For example, if a CFO can see that one supplier’s effective landed cost is rising faster than its peers, procurement can address that discrepancy before it compounds across future orders. Without this visibility, the margin hit is simply absorbed and explained later.
Similarly, inventory management becomes a strategic lever rather than a storage problem. While “front-loading” stock can hedge against future duty increases, it introduces massive working capital complexity. AI helps the CFO distinguish between genuine operational movement and tariff-driven shifts, clarifying which inventory batches reflect which cost structures. This improves decision-making around pricing and allows the CFO to weigh the actual exposure embedded in current stock against the carrying cost of additional inventory.
For CFOs in the US and UK facing immediate cost pressures, the transition to AI-enhanced monitoring should be an incremental evolution rather than a “big bang” transformation. Steinberg advises that the first step is identifying one specific area, a cost line or revenue stream where faster visibility would materially change a decision.
As AI takes over the “heavy lifting” of data examination, the technical barrier for finance teams is actually lowering. Modern tools are more intuitive, but the harder part is redefining what “good performance” looks like. The emphasis is shifting from data entry and reconciliation to commercial judgment.
Teams now need to understand how to assess the relevance of an alert and translate findings into strategic advice for the broader business. This requires a focus on “decision discipline” ensuring there is a clear route from an identified risk to an agreed action. If handled correctly, AI does not diminish the finance function; it gives it greater influence by allowing it to operate at a truly strategic level.