Sewage treatment plant workers in Portland raise concerns about artificial intelligence use
Artificial intelligence is showing up in places you would never expect — and not everyone is welcoming the change with open arms.
In Portland, Maine, workers at a wastewater treatment plant are at the center of a debate that blends technology, safety, and labor rights. The discussion is far from simple and goes well beyond a fight between humans and machines. What is really at stake is how technology enters critical environments where any mistake can put lives at risk and compromise essential services for thousands of people.
On Monday, members of Teamsters Local 340 gathered outside the Portland Water District office to bring their concerns directly to the board of trustees during a public meeting. The protest reflects a growing unease among workers at the East End Wastewater Treatment Plant, which serves nearly 15% of Maine residents across 11 communities, processing millions of gallons of sewage every year. That means the proper operation of this facility is not just an operational matter — it is a public health issue. Any failure in the process could have serious consequences for the environment and for the local population, which makes the debate around automation even more sensitive.
And that is exactly where the tension lies: rumors about adopting automation and AI in monitoring and treatment processes have raised a red flag among the workers there. It is not that the employees are against technology itself. The problem is the lack of clarity about how this transition would happen, which roles would be affected, and most importantly, whether safety standards will be maintained during and after the process.
The tension has already moved beyond the plant hallways and landed on the union bargaining table — with workers demanding more transparency and clear regulations before any changes take place. 🚨 This movement reflects a growing concern across multiple industries: the worry that the rush toward automation could steamroll risk assessment processes that exist specifically to protect both workers and the communities that depend on these services.
What the workers are saying
Wayne Dunphe, a treatment operator and Teamsters Local 340 member, summed up the team’s sentiment well when he said that nobody knows for sure whether automation is going to happen or when it might happen. According to him, the main problem is precisely that uncertainty. Rumors about the possible adoption of artificial intelligence for monitoring and treatment have been growing, but without any clear official communication from the district management.
For Dunphe and his colleagues, technology might help with certain processes, but it will never fully replace human judgment — especially in emergency situations. He was pretty straightforward in stating that if something goes seriously wrong, automation is not going to fix the problem. It is the human factor that makes the difference in those critical moments.
Dunphe’s concern is not abstract. He pointed out that human oversight is essential to ensure equipment is running properly and that bacteria levels and discharge standards stay within regulatory limits around the clock. In a hypothetical scenario where AI fails to detect a problem, sewage could end up being dumped directly into the bay — an environmental disaster that would affect the entire community.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
Joe Piccone, business agent for Teamsters Local 340, was even more emphatic. According to him, this is not just about rumors. Piccone stated that there have been concrete discussions about the possibility of eliminating human presence during night shifts, automating data measurement and discharge equipment. That prospect is concerning not only from an operational standpoint but also as a sign of a lack of investment in the local community.
Piccone made a point of reminding everyone that these workers also live in the area, drink the same water, and share the same environment. In other words, they have a personal stake in taking care of all of it. Removing these people from critical roles would not just be a technical decision — it would directly affect the commitment and quality of service provided to the public. 👷
What the Portland Water District says
On the other side of the table, the Portland Water District is trying to ease the tension. In an official statement, General Manager Scott Firmin said the district has no current plans to implement artificial intelligence in that way. Firmin acknowledged that there is a duty to ratepayers and customers to leverage technology to improve the workforce’s ability to operate efficiently, but he made clear that the district does not foresee any disruption to the workforce caused by automation in the near future.
The statement also mentioned that negotiations with the union are ongoing and that the Portland Water District continues to bargain in good faith, recognizing the value and importance of its workers.
Despite those assurances, union leaders remain skeptical. And honestly, that is understandable. The lack of a more detailed position on long-term plans leaves room for speculation, and in an environment where trust between management and workers needs to be solid, vague statements can do more damage than total silence.
Union members have already indicated that they will not agree to a new contract until there is more transparency about how AI could be used in plant operations. That firm stance shows that the issue is not just technical — it is a matter of trust and respect for the professionals who keep the system running day and night.
What AI has to offer in wastewater treatment
Before diving deeper into the conflicts, it is worth understanding what is at stake from a technological standpoint. Artificial intelligence applied to wastewater treatment is not science fiction — there are already systems being tested and deployed in different parts of the world with very promising results. Machine learning algorithms can analyze sensor data in real time, identify contamination patterns, and automatically adjust treatment processes with a level of precision that would be hard to achieve manually. This reduces chemical reagent waste, optimizes energy consumption, and can anticipate failures before they actually happen.
On top of that, AI-based systems can process massive volumes of data continuously, something human teams simply cannot do at the same speed and consistency. That does not mean replacing people but rather creating an additional monitoring layer that can make the process safer and more efficient. At large-scale plants like the East End, where the volume of processed sewage is enormous, this analytical capability can be a significant advantage in ensuring quality standards are always met.
However, all that technical capability only turns into a real benefit when implementation is done carefully. AI systems need to be trained with quality data, constantly monitored, and audited regularly. A poorly calibrated model in a critical environment like this one can make wrong decisions with very serious consequences. That is why the discussion about regulations is not bureaucracy — it is an essential part of responsible technology adoption. 🤖
Safety at the center of the debate
The most sensitive point in this entire discussion is safety. Wastewater treatment plants handle dangerous chemicals, high-pressure equipment, and processes that require constant monitoring. In environments like these, safety protocols are not suggestions — they are legal and operational requirements that exist to protect lives. When automation enters this context, the most important question is not how much money will be saved but rather how to ensure the level of safety will be maintained or, ideally, improved.
The East End workers are not wrong to ask for guarantees. Historically, poorly planned automation processes have already created dangerous situations in various industrial sectors, especially when implementation happens in a rush, without adequate training for the teams and without a detailed analysis of the specific risks in each environment. In wastewater treatment, where variables are numerous and the consequences of a failure can be severe, that caution makes complete sense and deserves to be taken seriously by municipal management and by the developers of the technological solutions.
Another important point is that artificial intelligence is still not infallible. AI systems make mistakes, especially when they encounter situations outside the dataset they were trained on. In a dynamic environment like a treatment plant, where weather conditions, sewage volume, and chemical composition can vary significantly, human oversight remains essential. The ideal approach, therefore, is not the complete replacement of workers with machines but rather a smart integration between human capabilities and the available technological tools. 🔍
The impact on night shifts is especially concerning
One of the most critical points raised by the union is the possibility of automating night shifts. According to Joe Piccone, discussions about eliminating human presence during the night have already taken place, and that is one of the biggest sources of concern among workers. Night shifts at treatment plants are not just a formality — they are periods when unexpected problems can arise with fewer resources available for a quick response.
Imagine a scenario where a piece of equipment fails at three in the morning and the automated system does not correctly identify the problem. Without any workers on-site to step in, sewage could be released without proper treatment, contaminating Casco Bay and putting the health of the entire community at risk. This type of situation is not hypothetical — similar incidents have already occurred at other facilities around the world when human oversight was reduced without the proper safeguards in place.
Having experienced operators present during all shifts functions as a safety net that no algorithm, no matter how sophisticated, can fully replicate at the current stage of the technology. Trained professionals can pick up on subtle anomalies — an unusual noise from a pump, a strange smell in the air, a vibration that feels off — that sensors might not catch. That combination of hands-on experience and intuition developed over years is something AI simply does not have yet.
Regulations: the path to responsible adoption
The movement by Portland workers points to something that goes beyond the specific case of that plant. There is a significant regulatory gap when it comes to the use of artificial intelligence in critical infrastructure. While sectors like healthcare and finance are already beginning to see more structured regulatory frameworks emerge for AI use, areas like sanitation and wastewater treatment are still in the early stages of that conversation. This creates a dangerous vacuum where high-impact decisions can be made without the backing of clear standards that guarantee safety for everyone involved.
In the American landscape, agencies like the EPA (Environmental Protection Agency) have the authority to set operational standards for treatment plants, but specific guidelines on the use of automation and AI in this context are still vague. That means, in practice, each municipality ends up making its own decisions without a clear guide on what is acceptable and what represents risk. For workers, that is a legitimate source of insecurity. For managers, it is an enormous responsibility that often does not come with the necessary tools to carry it out safely.
The most sensible solution involves building regulations that include all stakeholders: workers, managers, technology experts, environmental agencies, and representatives from the communities being served. This process needs to be transparent, evidence-based, and agile enough to keep pace with how quickly technology evolves. Automation has a lot to contribute to making wastewater treatment more efficient and sustainable — but that potential will only be realized in a positive way if implementation is carried out with responsibility, dialogue, and respect for the people who keep this system running every single day. 💡
A reflection of a global debate
What is happening in Portland is not an isolated case. Around the world, workers in multiple sectors are facing similar situations where the promise of efficiency brought by artificial intelligence runs into the reality of those who operate these systems on a daily basis. From the auto industry to logistics, through public services and infrastructure, the tension between automation and employment is one of the defining issues of our time.
What sets the Portland case apart is that it directly involves public health and the environment. Unlike a production line in a factory where a failure might mean financial loss, a failure at a sewage treatment plant could mean environmental contamination, disease, and long-term ecological damage. That dimension makes the debate even more urgent and reinforces the need for careful and inclusive approaches.
The East End workers are, in practice, asking for something quite reasonable: active participation in decisions that directly affect their work and community safety. This kind of dialogue between unions, management, and technology experts is not an obstacle to innovation — it is, in fact, what can ensure innovation happens in a sustainable and beneficial way for everyone.
The Portland case is an important reminder that technology and regulation need to move forward together. Rushing to adopt artificial intelligence solutions without proper attention to safety and clear standards can turn an opportunity into a problem — especially when it comes to essential services for the public.
