Agent-Based AI: What Packaging Manufacturers Need to Know
Eric Kutcher, President of McKinsey & Company for North America, put it better than anyone else: AI is “the true fourth industrial revolution,” as he said in a company podcast in January 2026. AI is as transformative as steam and steel, and it is spreading 100 times faster.
In December 2024, OpenAI surpassed 300 million weekly ChatGPT users. Earlier this year, that figure reached 900 million users per week. The most interesting statistic, however, is that 50 million of these weekly users are paying subscribers. And soon, OpenAI will launch its AI agents with the stated goal of strengthening its competitiveness against Anthropic AI.
According to McKinsey’s 2025 data, approximately 23% of companies are already evolving their use of agent-based AI. Nearly one-third of respondents worldwide have experimented with agents. But appearances can be deceiving: eight out of ten companies report that data limitations are a major obstacle to deploying agent-based AI at scale and deriving tangible value from it.
Understanding Horizontal vs. Vertical AI, and What "Agent-Based AI" Really Means
For packaging manufacturers, the adoption curve for agent-based AI is almost certainly closer to zero. The packaging sector is conservative, with specific requirements, and when it comes to adapting to a changing regulatory environment—which covers the food, pharmaceutical, and personal care industries—as with other demanding industries, this requires designing custom, zero-defect packaging.
Your sales, marketing, compliance, and executive teams have likely used ChatGPT, Copilot, and other general-purpose AI models for a variety of purposes: creating a presentation deck, prioritizing or monitoring emails, or conducting simple research. But general-purpose AI is what we call a cross-functional tool—it isn’t designed for a specific industry and doesn’t speak the same language as a packaging engineer with 20 years of experience.
In contrast, vertical AI designed for a specific industry can speak the same language as your packaging experts, understand the physics that determine how and why a package’s design must be structured in a certain way, and incorporate parameters specific to each customer. For a manufacturer of flat or corrugated cardboard, it can draw on the expertise of a packaging engineer with 20 years of experience, provided it has been trained to do so.
That's where agents come in. The difference between a chatbot and an agent is like the difference between an estimator on their first day and a senior estimator whom you trust to manage an entire account.
The chatbot is like a first-day estimator: it does exactly what you ask it to do—and nothing else. Ask it for a price on a specific item, and it’ll give it to you. But it won’t open the file on its own, won’t notice that the requested quantity doesn’t make sense, and won’t check if the box is even available. Unless you ask the exact question, nothing happens.
The agent is the senior estimator to whom you assign a client brief. He reviews the specifications, verifies whether the project is feasible, notes that cardboard prices have changed since the last quote, makes adjustments, and comes back with a complete proposal. You set a goal for them, not a list of steps. And the more projects they handle, the more refined they become—because they learn to adapt.
But AI is also subject to rules that it must follow. An agent will know when it’s best to seek a human decision or validation. Crucially, agents do not have the authority to make decisions on your behalf—they are trained and supervised to do exactly what a manufacturer needs at any given time. Pricing, processes, packaging performance, and more: control and decision-making authority remain in the hands of the manufacturer.
How to Use Agent-Based AI and Expand Its Use
As a leader in the flat or corrugated cardboard industry, effectively using agent-based AI requires learning by doing. When a customer reaches out to you, they are looking for a specific outcome. Today, 40% of the data an agent needs resides in the knowledge of your packaging experts—not in your structured systems, such as your ERP or MIS.
Our simple guide can help you rethink the way you think:
1. Rethink Your Processes with AI
This is the most important—and most time-consuming—part of the work. It involves identifying where knowledge resides within your company and how to effectively extract it for use by AI. This may involve working with your sales representative who manages several key accounts and has specific information about each client, your engineers who know packaging formats and specifications like the back of their hands, and your compliance and legal managers who are familiar with regulatory requirements—all integrated into an AI training model.
2. Audit these processes and protect your data
Use this insight to identify the areas that would benefit most from an AI agent. For example, fluctuations in cardboard prices may have affected your pricing model. Do you need an agent capable of autonomously detecting price inconsistencies based on your quote history? It could provide data-driven insights to optimize your pricing, then hand things off to your sales team for the final decision. It’s by identifying these specific moments that an agent can make a real difference.
3. Start small, grow fast
Don’t run before you can walk. Focus on a specific use case and measure its return on investment before scaling up. Executive teams reap the greatest benefits from AI when they avoid “big bang” transformation programs.
4. Think about your carbon footprint
AI consumes a lot of energy and scarce resources. Choosing a vendor that can track and report the carbon footprint of its AI usage is a best practice for any company.
How Packitoo Can Help You
At Packitoo, we develop a range of AI agents designed to support your teams. They are built with specialized capabilities based on both your structured data (ERP/MIS) and your unstructured data (valuable insights and knowledge specific to your customers and packaging).
Experts in flat and corrugated cardboard can join an exclusive group of members to gain access to our AI agent platform. Currently in beta, more information about this platform will be revealed in an upcoming blog post.
In the meantime, packaging industry professionals have access to HIPE. Our cloud-based digital sales platform helps you optimize your product configuration, pricing, and quote generation (CPQ) processes, and turns these time and performance gains into tangible benefits. Combined with our AI agent program, it can help you automate your sales tasks, uncover untapped opportunities in your pipeline, meet customer deadlines, increase your margins, and provide support for technical questions. By choosing Packitoo, you’re partnering with a software provider that not only knows your industry inside and out but can also guide you as we enter this fourth industrial revolution. To move forward step by step, contact our team of AI experts today.