Are Forced Labor and Tariff Compliance One AI Planning Problem?

The hard procurement question in 2026 is no longer whether forced-labor compliance and tariffs both matter. It is whether an AI supply chain planning system can recommend, compare, and defend a sourcing move when one supplier reduces tariff exposure and another introduces forced-labor risk somewhere below tier one.

That question lands badly in most organizations because the two clocks run in different rooms. Tariff scenarios move through S&OP cycles, landed-cost models, sourcing events, and executive trade-off meetings. UFLPA enforcement moves through shipment exams, entity-list screening, origin documentation, and rebuttal windows. The planner is asked for a fast answer. The compliance team is asked for a defensible one. When the answer is not the same, somebody inherits the reconciliation work.

The collision is now large enough to treat as a planning architecture problem. DHS reported more than 16,700 shipments examined under UFLPA enforcement, more than 10,000 denied shipments, roughly $3.7 billion in examined value, and 144 entities on the UFLPA Entity List in its August 2025 strategy update; CBP’s enforcement dashboard remains the live source for current shipment counts.[1][2] On the tariff side, Kinaxis said tariff debates had triggered a 124% spike in scenario-planning activity and a 4.5x increase in daily activity among auto-sector customers in March 2025, while launching a Tariff Response solution with a stated 21-day deployment timeline.[3]

Those figures do not prove that any one platform solves the combined problem. They do show why separating the work into “trade compliance” and “tariff planning” creates a false handoff. A sourcing recommendation is not usable if the assumed supplier path cannot clear origin scrutiny. A compliant supplier choice is not complete if it ignores the tariff exposure that triggered the sourcing event in the first place.

What “One AI Planning Problem” Actually Requires

A true dual-risk planning engine would not merely display tariff alerts beside forced-labor alerts. It would let compliance facts govern the feasible solution space before the optimization result is presented to planners.

That means blocked entity lists, XUAR-origin restrictions, multi-tier traceability gaps, documentation status, and CBP rebuttal timing must be represented as constraints or decision states inside the same environment that models tariff cost, lead time, capacity, supplier qualification, inventory position, and customer-service impact. If a supplier route cannot be defended in an enforcement window, the planning engine should not treat it as an ordinary cost-saving option and leave compliance to clean up the exception later.

This is where the market language gets slippery. A risk graph can identify exposure. A document workflow can collect supplier attestations. A tariff agent can scan duty changes. A planning optimizer can rebalance demand, supply, inventory, and constraints. The buying question is whether those capabilities share decision authority, not whether they appear in the same sales deck.

The practical standard is simple: when a planner runs a tariff scenario, the system should know which supplier moves are legally or evidentially unavailable, uncertain, or time-bound. If it only warns the planner after the optimization run, it is not yet solving the same problem.

Why Lower-Tier Exposure Keeps Breaking Clean Scenarios

Direct import shifts can make dashboards look cleaner while lower-tier risk remains. CSIS assessed in August 2025 that Xinjiang’s global polysilicon share fell from about 41% to 24.8%, an important movement in a sector heavily associated with UFLPA scrutiny.[4] RAND’s January 2025 research made the complementary point: direct XUAR imports declined, but exposure can persist through tier-2, tier-3, and tier-4 supplier relationships.[5]

That mechanism matters more to planning teams than the headline location of a tier-one supplier. A component sourced from a lower-tariff country may still rely on upstream inputs, labor, or entities that create detention risk. Conversely, a higher-tariff option may be the only supplier route with documentation strong enough to ship without a foreseeable compliance interruption. The “best” plan depends on which constraint is allowed to disqualify the scenario.

For readers working through the traceability side of the problem, ChainSignal’s analysis of Altana for trade compliance is a useful adjacent reference. It sits closer to the compliance-tracing dimension than to native tariff optimization, which is exactly the distinction procurement teams need to preserve.

How the Major Platforms Read Against the Dual-Risk Test

The current platform field is not empty. It is uneven. Each vendor has a credible part of the answer, and each part matters. The gap appears when a buyer asks whether forced-labor constraints and tariff scenarios are being resolved by one decision engine or passed between adjacent modules.

Kinaxis: Strongest Signal on Tariff Response and Scenario Execution

Kinaxis is easiest to read as a tariff-response planning story. Its April 2025 Tariff Response release emphasized rapid deployment, automated tariff impact analysis, scenario evaluation, and the ability to rebalance supply plans under disruption. The vendor’s own activity figures — the 124% scenario-planning spike and 4.5x auto-sector daily activity increase — are useful because they show what customers were doing when tariff volatility hit, but they remain vendor-published usage data rather than neutral evidence of planning effectiveness.[3]

For a planning director, the Kinaxis strength is not that it says “tariff AI.” The strength is that tariff response is close to the planning motion: scenarios, constraints, alternatives, and execution timing. That is where tariff volatility belongs. The unresolved question is whether UFLPA-style constraints — entity exposure, XUAR-origin evidence, documentation sufficiency, and rebuttal timing — can actively remove or downgrade sourcing options inside the same planning run, rather than arrive through another compliance system.

If the tariff scenario engine recommends a supplier move that later requires manual compliance intervention, the downstream work has not disappeared. It has merely moved from optimization into exception management.

o9: Stronger on Connected Modeling, Less Clear on Compliance as a Hard Constraint

o9’s June 2025 release is closer to the architecture many enterprises say they want: an Enterprise Knowledge Graph for multi-tier modeling, planning and collaboration capabilities for reconfiguring global supply chains, supplier relationship management for compliance documentation, and tariff what-if scenarios.[6] That combination matters because lower-tier supplier facts are not useful if they cannot be connected to demand, supply, and sourcing choices.

The positive reading is that o9 has the modeling language for the combined problem. Multi-tier relationships, supplier collaboration, and planning scenarios can live in a connected environment instead of being rebuilt in spreadsheets. The procurement caveat is the same one that should appear in any evaluation: documentation-adjacent planning is not identical to compliance-constrained optimization. A supplier file can be present, incomplete, expired, disputed, or insufficient for rebuttal. The planning engine needs to treat those states differently.

o9 therefore looks strongest where the enterprise already has the organizational discipline to maintain supplier graphs and compliance documentation. Without that discipline, the knowledge graph becomes a better map of uncertainty, not automatically a better decision.

Blue Yonder: Network Visibility Plus Tariff Sensing

Blue Yonder’s April 2025 tariff discussion emphasizes a digital supply chain advantage built around network-based multi-tier visibility and a Tariff Agent that scans real-time tariff data.[7] That is a practical pairing. Tariff changes are only actionable if the system knows which suppliers, lanes, materials, and customers are touched.

The limitation is in the word “sensing.” Sensing a tariff change is not the same as solving the constrained sourcing problem, and visibility into a network is not the same as proving forced-labor admissibility. Blue Yonder’s position is relevant for teams that need tariff awareness distributed across a supply network, especially where the first failure mode is slow detection. It is less conclusive on the question of whether forced-labor compliance constraints govern the same optimization run that evaluates tariff alternatives.

Resilinc: Explicitly Puts Forced Labor and Tariffs Behind One Interface

Resilinc is unusually direct in naming both sides. Its October 2025 agent framing includes a Forced Labor Compliance Agent and a Tariffs Agent accessed through a single conversational interface, supported by what the company describes as a 15-year validated supplier graph.[8] For teams drowning in alerts and supplier data, that interface choice is not cosmetic. It acknowledges that the user asking about tariffs may also need to know whether the supplier route can survive forced-labor scrutiny.

The evaluation issue is whether the agent is informing a planner or constraining a plan. A conversational interface can reduce search time, summarize exposure, and route the next action. It does not, by itself, prove that tariff optimization, supplier feasibility, compliance status, and execution timing are being solved together. Resilinc’s framing is valuable because it reflects the way procurement questions now arrive. It still needs to be tested at the handoff from insight to decision.

Interos: A Risk-Intelligence Lens, Not a Native Planning Optimizer

Interos is better understood as a risk-intelligence lens than as the planning optimizer itself. Its forced-labor and tariff commentary estimates that 1.3 million companies are at high risk for unethical labor, with nearly half directly supplying U.S. companies.[9] That kind of exposure estimate is useful for prioritization, especially when an enterprise needs to understand where risk may be hiding beyond its direct supplier base.

The planning gap is familiar. Knowing that a supplier network carries elevated risk does not automatically answer which alternate supplier, inventory posture, production timing, or customer allocation should be selected under tariff pressure. Interos can sharpen the risk signal that planning should respect. It is not, on the evidence available here, the native engine that resolves the complete dual-risk optimization problem.

The Vendor ROI Claims Are Useful, but They Are Not the Decision Standard

FRDM’s AstraZeneca case is a good example of why supplier-depth analytics attract procurement attention. FRDM reports that AstraZeneca mapped supply chains six tiers deep, found that 8 entities drove more than 90% of warnings, and achieved a 94.6% due-diligence cost reduction; those are vendor-reported figures, not independently verified benchmarks in this research set.[10]

The case still has operational value because it points to a real mechanism: risk is often concentrated, and mapping deeper tiers can change where teams spend investigative time. But due-diligence efficiency is not the same as optimized tariff response. A platform can reduce compliance review cost and still leave planners to rebuild the sourcing recommendation in another system.

That is the procurement trap in many demos. The compliance demo proves that the buyer can see more. The planning demo proves that the buyer can simulate more. The missing proof is that the system refuses, penalizes, or time-bounds a tariff-attractive move when the forced-labor evidence is insufficient.

What Buyers Should Test in 2026

The useful RFP question is not “Do you have tariff AI?” or “Do you support forced-labor compliance?” Almost every serious vendor can say something credible to both. The better question is: when a tariff what-if scenario identifies a lower-cost supplier path, can compliance constraints actively govern whether that path is feasible, conditional, or blocked?

• Ask the vendor to run one scenario where the lowest-cost tariff option has a lower-tier origin problem, then show whether the planning result changes before manual review.
• Require separate labels for tariff sensing, risk visibility, documentation workflow, and constrained optimization; do not accept one label covering all four.
• Test whether entity-list updates, origin restrictions, and documentation status change supplier feasibility inside the planning model.
• Trace who owns the exception when the system recommends a sourcing move that compliance cannot defend.
• Check whether the audit trail explains why an option was excluded, not just which option was selected.

The EU Forced Labor Regulation adds another deadline pressure point for platform selection. FRDM and Interos both discuss enforcement beginning in December 2027, although this research pass did not independently verify that date against the official EU legal text.[9][10] For enterprises choosing systems in 2026 and 2027, the timing matters less as a legal interpretation than as an implementation constraint: supplier mapping, documentation workflows, planning integrations, and audit trails will not mature overnight.

For tariff-specific planning depth, ChainSignal’s related analyses on AI tariff scenario planning and planning for permanent tariffs are the cleaner places to separate scenario mechanics from the compliance constraint problem.

The Market Gap

As of Q3 2026, the market offers strong partial systems. Kinaxis is strongest in the evidence reviewed here on tariff response and scenario execution. o9 is strongest on connected multi-tier modeling and planning-adjacent supplier documentation. Blue Yonder is strongest on network visibility plus tariff sensing. Resilinc is unusually explicit in placing forced labor and tariffs behind one conversational interface. Interos brings a risk-intelligence view that can inform planning but should not be mistaken for native optimization.

None of the reviewed evidence establishes a native end-to-end engine that enforces forced-labor compliance constraints while running tariff what-if scenarios in one optimization environment. That does not make the platforms weak. It means the procurement decision is about architecture fit, integration burden, auditability, and accountability.

The team evaluating AI supply chain planning for forced labor and tariffs should therefore follow the exception. If the platform recommends a supplier move that enforcement reality later blocks, who sees the conflict first, which system changes the plan, and who can explain the decision before the shipment clock runs out?

References

1. 2025 Updates to the Strategy to Prevent the Importation of Goods Mined, Produced, or Manufactured with Forced Labor in the People’s Republic of China, DHS, August 2025, https://www.dhs.gov/2025-updates-strategy-prevent-importation-goods-mined-produced-or-manufactured-forced-labor-peoples
2. Uyghur Forced Labor Prevention Act Enforcement Statistics, CBP, https://www.cbp.gov/newsroom/stats/trade/uyghur-forced-labor-prevention-act-enforcement-statistics
3. Kinaxis Launches Tariff Response Solution to Help Supply Chains Adapt to Disruption with Confidence, Kinaxis, April 2025, https://investors.kinaxis.com/news-releases/news-release-details/2025/Kinaxis-Launches-Tariff-Response-Solution-to-Help-Supply-Chains-Adapt-to-Disruption-with-Confidence/default.aspx
4. Assessing the Impact of the Uyghur Forced Labor Prevention Act After Three Years, CSIS, August 2025, https://www.csis.org/analysis/assessing-impact-uyghur-forced-labor-prevention-act-after-three-years
5. Research Report RRA2534-1, RAND, January 2025, https://www.rand.org/pubs/research_reports/RRA2534-1.html
6. o9’s Comprehensive Planning and Collaboration Capabilities Help Enterprises Rapidly Reconfigure Their Global Supply Chains, o9 Solutions, June 2025, https://o9solutions.com/news/o9s-comprehensive-planning-and-collaboration-capabilities-help-enterprises-rapidly-reconfigure-their-global-supply-chains
7. Mitigating Tariff Impacts: The Digital Supply Chain Advantage, Blue Yonder, April 2025, https://blueyonder.com/blog/2025/mitigating-tariff-impacts-the-digital-supply-chain-advantage
8. Supply Chain Teams Are Drowning in Data. AI Agents Throw the Lifeline, Resilinc, October 2025, https://resilinc.ai/blog/supply-chain-teams-are-drowning-in-data-ai-agents-throw-the-lifeline/
9. Supply Chains Under Pressure: Forced Labor, Shifting Tariffs, and the China Summit, Interos, https://www.interos.ai/blog/supply-chains-under-pressure-forced-labor-shifting-tariffs-and-the-china-summit
10. Can Artificial Intelligence Fix Slavery in Supply Chains?, FRDM, https://www.frdm.ai/resources/can-artificial-intelligence-fix-slavery-in-supply-chains