Amazon Australia is scaling its packaging reduction strategy with the help of artificial intelligence, using a new tool that determines whether products can be shipped in their original retail packaging without the need for an added delivery box or bag. Developed initially in Australia with support from local AWS teams, the system marks an important step in how e-commerce players are using data and automation to cut waste while maintaining product protection across large fulfilment networks.

The AI model analyses product information to assess whether an item is robust enough to travel through Amazon’s logistics system in its own packaging. This replaces much of the time-consuming manual review previously required to decide if a product qualified for shipment without extra outer packaging. According to Amazon, the impact on speed has been significant: the company was able to add 12,000 products to the programme in just one month, compared with almost 18 months under the previous manual testing process.

That acceleration has expanded the number of eligible items in Australia to more than 150,000 products, showing how automation can unlock scale in sustainable packaging decisions. For the packaging industry, the development is notable because it shifts packaging optimisation from a static design exercise to a more dynamic, data-led process. Instead of relying only on standard packaging rules, Amazon is using machine learning to decide where packaging can be eliminated altogether and where it remains essential for damage prevention.

The initiative forms part of a broader push to reduce waste across Amazon’s Australian operations. The company says that one in 10 orders shipped from an Australian fulfilment centre is now delivered in its original packaging, while the number of orders sent to customers with no added delivery packaging has more than tripled since 2021. Amazon has also phased out single-use plastic delivery bags in Australia in most cases, replacing them with flexible paper bags, padded envelopes, paper dunnage and water-activated tape that can be recycled through kerbside systems.

For online retail, this reflects a wider market shift. Consumers increasingly associate oversized boxes, excessive void fill and unnecessary secondary packaging with wasteful e-commerce practices. As a result, reducing packaging has become both an environmental objective and a customer experience issue. In this context, Amazon’s use of AI is not simply about cutting material usage; it is also about improving packaging accuracy so each product receives only the level of protection it actually needs.

The most sustainable e-commerce package is often no added package at all, but reaching that point at scale requires better data, faster decisions and confidence that products can still arrive intact.

Amazon says that globally 50% of customer orders now ship with less packaging or no added packaging, and that its packaging reduction efforts have avoided more than four million metric tonnes of packaging since 2015. The Australian rollout shows how regional teams can play a leading role in that strategy, particularly when digital tools are adapted to local logistics and recycling conditions.

For Packnode readers, the significance of the story lies in the growing connection between AI, packaging optimisation and e-commerce efficiency. As fulfilment networks become more automated, packaging decisions are also becoming smarter and more precise. Amazon Australia’s programme suggests that the next phase of sustainable e-commerce packaging will be driven not only by new materials, but by intelligent systems capable of reducing packaging at speed and scale without compromising performance.