Extended producer responsibility, or EPR, is becoming a critical compliance issue for companies selling packaged goods in the United States. As more US States introduce their own packaging rules, businesses need to understand how much packaging they place on the market, how that packaging is classified by these programs, where their consumers are discarding it, which exemptions apply, and what fees they’re obligated to pay. For companies with thousands of SKUs, multiple packaging formats, and incomplete historical data, the operational burden of EPR compliance can be substantial and is orders of magnitude larger than it has been in years past.

Net Zero Compare spoke with Arjun Gupta, Founder and CEO of MokshEPR, about how packaging EPR works in practice, why US requirements are becoming more detailed, where companies run into data problems, and how AI can support compliance without removing the need for human review.

🎥 Watch the Full Conversation: The full interview goes deeper into the operational side of packaging EPR, including examples of how state-level rules affect reporting and how packaging choices can influence future costs. Arjun also explains how MokshEPR approaches AI-assisted classification, data cleaning, forecasting, and human review. The conversation provides additional context around the trade-offs companies face when building an EPR compliance process.

Why packaging EPR is becoming more complex

At its core, EPR shifts more of the financial responsibility for packaging waste toward the companies that place that packaging on the market. Arjun described the principle as linking the cost of collection, sorting, and recycling more directly to producers. The difficulty is that reporting increasingly goes beyond broad categories such as plastic, paper, glass, or metal and requires a much more detailed understanding of individual packaging components.

A single product may include a bottle, cap, label, outer carton, tray, and secondary packaging, with each component requiring separate evaluation. Classification may depend not only on the material itself, but also on how the packaging is used, its form factor, whether it stays attached to another component when discarded, and whether it is already covered by another recycling program. Arjun gave the example of a label attached to a bottle, which may be treated together with the bottle depending on the applicable methodology because both are likely to move through the same waste stream. He also pointed to small packaging components that can create different sorting challenges because of their physical size. All of these factors impact the cost of collection, recycling, and in the end – the cost of circularity of the system as a whole. The categorization & fees for different packaging in any state are reverse-engineered from these cost drivers.

These requirements force companies to track operational details that may never previously have been captured in a structured way. For businesses with large product portfolios, the challenge is therefore not simply understanding the regulation but also translating existing product and packaging information into the level of detail required for reporting.

State-by-state rules add another layer of complexity

For businesses selling across the US, one of the main challenges is that reporting methodology is not identical from state to state. Arjun used bottle deposit programs as an example. Certain packaging may already fall under a separate bottle bill or redemption program, and if it does, it may be treated differently for EPR purposes. The exact treatment, however, depends on the state.

As a result, a company cannot assume that a methodology developed for one state can simply be reused in another. The same SKU may require different treatment depending on where it is sold, how the state defines covered materials, which exemptions apply, and how other waste programs interact with EPR. In practice, compliance becomes less about producing one national packaging report and more about applying different regulatory logic to the same underlying product data.

That distinction matters for companies trying to standardize internal reporting. A single centralized packaging dataset may still be useful, but the rules applied to that data can vary by jurisdiction, which means companies need both consistency in the underlying information and flexibility in how it is interpreted.

The real bottleneck is often packaging data

Arjun repeatedly pointed to data quality as one of the main practical problems companies face. Packaging information is often spread across procurement systems, supplier spreadsheets, product records, logistics files, and internal documents. Some fields may be missing altogether, including packaging weight, dimensions, material composition, color, format, or the relationship between specific components and SKUs.

According to Arjun, MokshEPR starts by normalizing this information into a structure that can be reviewed more consistently. The platform then uses AI to help identify gaps, pre-fill certain decisions, and organize the data required for different reporting methodologies. The objective is not to replace internal expertise, but to reduce the amount of repetitive processing required before a human reviews the result.

That distinction is central to Arjun's approach. A software system may understand the reporting methodology, but the company itself is more likely to know whether a particular packaging assumption is correct. Packaging, sustainability, procurement, and other internal teams therefore remain part of the process, especially when the available data is incomplete or requires interpretation.

AI can assist decisions without replacing accountability

A recurring theme in the conversation was the difference between using AI to assist a decision and relying on it to make an unreviewed compliance decision. Arjun said MokshEPR uses AI to pre-fill classifications and other decisions while providing reasoning and confidence indicators for review. The aim is to automate the more straightforward parts of the work and direct human attention toward unclear or incomplete data.

That matters in a regulatory context because companies remain responsible for the information they report. AI can speed up classification, data processing, and review, but it does not remove the need to validate assumptions, retain supporting evidence, or understand how a particular conclusion was reached. Human oversight is therefore not an optional safeguard added at the end of the process. It is part of the workflow itself.

Arjun described traceability as another important element. According to him, MokshEPR allows users to review the decision path behind a SKU, including who approved or changed particular values and what reasoning supported an AI-generated result. For compliance teams, this kind of audit trail can be as useful as the final report because it helps explain how reported figures were produced and where human judgment was applied.

Better data can materially affect fees

EPR reporting is not only an administrative requirement. It can have a direct financial effect because even small changes in classification can become material once they are multiplied across large sales volumes. If a company sells hundreds of thousands of units, differences in the classification, weight, or material type of a packaging component can change the overall fee calculation.

The same applies when different materials carry different fee levels. Arjun noted that plastics can attract substantially higher costs than some other materials because fee structures are intended, in part, to reflect recycling costs and encourage shifts toward more recyclable packaging. For large producers, small differences at the component level can therefore add up quickly across an entire product portfolio.

This is why accurate packaging data matters beyond filing. If a company does not understand its underlying data, it may struggle to forecast future EPR costs or identify which packaging components are driving the largest share of its fees. Better data can therefore support both compliance and financial planning.

Forecasting may become as important as filing

Arjun argued that companies should not treat EPR as a once-a-year reporting exercise. Once packaging data has been structured properly, it can also support planning and scenario analysis. A company can begin evaluating what would happen if a plastic bottle were replaced by another format, how that change would affect EPR fees, what the production change would cost, how labels or secondary packaging might be affected, and when the investment might pay for itself.

MokshEPR has developed forecasting functionality around this type of analysis. Arjun described it as a way to compare current packaging with potential alternatives and estimate how those decisions could affect fees and packaging waste over time. The value of this approach is that packaging decisions can be assessed not only in terms of immediate procurement costs, but also in terms of longer-term regulatory and operational consequences.

The broader implication is that reliable EPR data can support procurement, packaging design, cost planning, and investment decisions, not just regulatory reporting. Once the data exists in a structured form, companies can use it to evaluate changes before those changes are implemented.

Packaging data can support broader operational decisions

Arjun pointed to packaging redesign, transportation efficiency, right-sizing, supplier decisions, and waste reduction as areas where detailed packaging data can become useful. One example discussed in the interview involved the difference between round bottles and square packaging formats. A change in shape can affect how much product fits into a truck, which in turn can influence logistics efficiency as well as packaging-related costs.

Another example was right-sizing. A package may technically comply with reporting rules but still contain more material or space than necessary. If a company can identify products where packaging can reasonably be reduced, it may be able to lower material use while also reducing future fee exposure. This type of analysis requires more than a high-level material category. It depends on understanding dimensions, weights, component relationships, and the effect of those choices across a product portfolio.

This is where structured packaging data becomes more useful operationally. Instead of treating packaging only as a procurement expense or reporting input, companies can evaluate it as part of a wider system that affects logistics, cost, waste, and sustainability performance.

Lower EPR fees and lower environmental impact are related, but not identical

Arjun argued that EPR systems are designed to connect financial incentives with environmental outcomes by attaching costs to packaging choices. That can help sustainability teams and finance teams work toward overlapping goals, because reducing material use or changing packaging formats may also reduce future fee exposure.

The relationship is not always perfectly aligned, however. Some companies may already use very lightweight or highly optimized packaging and still face unfavorable treatment under a particular fee structure. Arjun gave flexible packaging as an example. A company may already have reduced plastic to the minimum needed for product protection, yet still face higher costs because the material itself is difficult to recycle.

This highlights one limitation of the current system. EPR methodologies will continue to evolve as regulators and producer responsibility organizations collect more information about collection, sorting, and recycling outcomes. Companies should therefore avoid treating today's reporting framework as a fixed end state and should expect methodologies, classifications, and incentives to change as the programs mature.

EPR requires input from several teams

Another practical issue is internal ownership. Arjun said companies often assign EPR responsibility to one or two people in sustainability or compliance, even though the information needed for reporting sits across several functions. Packaging teams may know the bill of materials, sales teams may know where products were sold, logistics teams may understand pallets, shrink wrap, and secondary packaging, while sustainability and compliance teams are more familiar with the reporting methodology.

Finance teams also have a role because EPR fees affect forecasting and cost planning. For larger companies in particular, EPR is therefore better treated as a cross-functional workflow than as a narrow reporting task. The person responsible for filing may coordinate the process, but that does not mean all of the underlying work should sit with that individual.

Clear ownership of different data inputs can reduce bottlenecks and improve accuracy. It also makes the process more sustainable over time because the information is gathered by the teams that are closest to it rather than repeatedly reconstructed by a central compliance function.

Avoid overprocessing the data too early

Arjun also warned against doing too much manual normalization before companies understand what is actually required. One example is breaking shared logistics materials, such as pallets, into tiny fractions across individual SKUs. That may appear to make the data more precise, but repeated transformations can make it harder to trace how the final figures were produced.

The same problem can arise when companies combine several SKUs into simplified reporting structures purely to reduce the number of calculations. Those shortcuts may make the initial filing process easier, but they can reduce transparency and make future forecasting more difficult. If the source data has been transformed repeatedly, it becomes harder to understand how a final number relates back to the original operational information.

Arjun's view is that companies should prioritize confidence and traceability over creating a heavily manipulated dataset. As automation improves, there is less reason to simplify data purely because a human would otherwise need to process every line manually.

What companies should do first

For companies that have barely started preparing, Arjun recommended beginning with the underlying packaging data rather than immediately focusing on software, consultants, or fee optimization. The first step is to understand what information already exists, what is missing, and which SKUs may be in scope.

That can include requesting packaging specifications from suppliers and mapping the data required in the jurisdictions where the company sells products. Suppliers are also becoming more familiar with these requests as more customers ask for the same information, which can make the data collection process more manageable than companies initially expect.

Once the basic data is assembled, companies can begin cleaning it, validating it, and identifying the most important gaps. Only then does it make sense to build reliable reports, compare compliance options, or evaluate packaging changes. Starting with the data also gives companies a clearer understanding of the scale of the work before they commit significant time or money to a particular solution.

What to expect over the next few years

Arjun expects US packaging EPR reporting to become more detailed rather than less. He also expects the volume of work to increase as more states move from simplified reporting toward more granular methodologies. At the same time, similar changes are taking place outside the US, creating additional complexity for businesses that operate across multiple markets.

MokshEPR is currently focused primarily on the US market while also responding to customer demand in Canada and selectively supporting other jurisdictions. For multinational businesses, the challenge is that packaging compliance is becoming more fragmented at the regulatory level while companies are still trying to manage packaging data across global product portfolios.

That makes structured, reusable packaging data increasingly valuable. Companies that can maintain one reliable source of packaging information and apply different regulatory methodologies to it are less likely to rebuild the process from scratch every time a new rule appears.

Conclusion

Packaging EPR is becoming as much a data and operational issue as a regulatory one. Companies need reliable information on what packaging they use, how it is classified, where it is sold, which exemptions apply, and how those factors affect both reporting and future costs.

AI can reduce some of the manual work involved in classification, data cleaning, and reporting, but human review remains essential because internal teams still hold the context needed to validate assumptions and take responsibility for the final result. The most practical starting point is therefore to get the packaging data in order, identify the gaps, assign responsibility across the relevant teams, and preserve enough traceability to support both compliance and future business decisions.