The entry of generative artificial intelligence into the drafting of technical offers is transforming the preparation of tenders, raising complex questions about the operational stability of public construction sites. If on the one hand the software guarantees detailed and visually impeccable formal reports, on the other hand there is a real risk of separating documentary perfection from the real executive capacity of the company in the field.
Technical offers and tenders: the risk of a distance between paper and construction site
Mauro Ferrarini: President Corbo, first of all let’s address a central issue for the future of public works. The new Public Contracts Code (Legislative Decree 36/2023) focuses on the “Principle of result” (Art. 1). How does the abuse of technical offers drawn up by artificial intelligence risk betraying this principle in the executive phase of the construction site?
Roberto Corbo: The first risk is being impressed by the apparent perfection of a technical offer. Artificial intelligence is capable of producing extremely detailed reports, complete in every respect, which can generate great expectations. But who is really behind that proposal? When you move from paper to construction site, it is people who create the works, with their skills, manual skills, experience and ability to deal with unexpected events. It is not an artificial intelligence that enters the construction site and carries out the work. The danger is creating a fracture between a virtual world, in which everything appears perfect, and the real one, made up of concrete problems, responsibilities and operational solutions. The principle of results requires us to look precisely at this second aspect: it is not enough to present an impeccable project, we must demonstrate that we are capable of realizing it.
Mauro Ferrarini: You underlined that the form of a proposal cannot exceed the value of the substance and constructive experience. Concretely, which criteria for evaluating the most economically advantageous offer (OEPV) should contracting authorities update or enhance to “shield” tenders from purely algorithmic relationships?
Roberto Corbo: I wouldn’t talk about races that need to be locked down, because we must always guarantee maximum transparency and the widest possible participation. However, I believe that it is necessary to give greater value to the history and reliability of companies. A first intervention could concern the participation of consortia in tenders below certain economic thresholds. We are talking about subjects who, in some cases, do not carry out the work directly, but make their requirements available to companies that may not have adequate experience and structure. It is a mechanism on which I believe it is appropriate to open a reflection.
The other element is to introduce rewarding criteria linked to the results achieved: work completed on time, business continuity, reliability demonstrated over the years and ability to keep commitments. Let’s imagine a company that has been active since 1950, which has gone through decades of economic transformations, has remained on the market and has always respected its obligations. This wealth of experience must have a recognizable value, we cannot put on the same level those who have built a reputation over time and those who present themselves on the market without a history behind them, perhaps aiming exclusively to win as many orders as possible. Competition must remain open, but it must also reward the demonstrated ability to complete the work.
Opportunities and risks for businesses
Mauro Ferrarini: The Code strongly encourages the digitalisation of the contract life cycle and the use of BIM. How do you reconcile your call for caution towards AI with the necessary digital transition that structured construction companies are called upon to face? Where do you draw the line between virtuous innovation and “dangerous shortcut”?
Roberto Corbo: I’m not against artificial intelligence, far from it. I believe it represents an extraordinary opportunity for businesses, especially in the organization of work, process management, planning and preventive identification of critical issues. It can help us analyze data, improve forecasts, optimize time and resources. The digital transition is necessary and companies must invest in this direction.
The boundary, however, is very clear: AI must be a tool to support human skills, not a system called upon to completely replace them. A construction company is not only made up of documents, processes and procedures, but of professionals, technicians, workers and people who take decisions and responsibilities every day. Technology can make them more efficient, but it cannot replace the experience gained in the field. Innovation is virtuous when it strengthens an already competent organization; it becomes a dangerous shortcut when it is used to mask skills that don’t exist.
Mauro Ferrarini: However, there is another point of view: for many small and medium-sized enterprises (SMEs), historically disadvantaged in the departments of tender management compared to large groups, generative AI represents a tool of democratization. Don’t you think that limiting its use in tender criteria could, in fact, penalize competition and protect only already established players?
Roberto Corbo: I also see the opposite risk. A structured company has engineers, architects, professionals and internal skills that can use artificial intelligence to improve their work. We, as a company, are also investing in employee training to integrate these tools into company processes. Our goal is to enhance people, not replace them.
Instead, I fear that it is precisely the smaller and less structured companies that are most exposed to superficial use of technology. Today everyone talks about artificial intelligence, just as in the past everyone talked about Superbonus or PNRR. And around every new opportunity a market of miracle solutions inevitably develops. There are entities that promise entrepreneurs that they can do without employees, technicians or administrative staff thanks to a computer program. It is a very dangerous message, especially for those who do not have the tools necessary to evaluate it. I believe that true democratization does not consist in replacing skills with an algorithm, but in also enabling small businesses to use these technologies in an informed manner.
The challenge of reconciling simplification and reliability
Mauro Ferrarini: Let’s put ourselves in the shoes of a RUP (Sole Project Manager) who has to evaluate dozens of offers: distinguishing a report drawn up by an AI from one written by a pool of senior engineers is now almost impossible for tender commissions. Doesn’t your proposal to give more weight to “organizational solidity” risk burdening bureaucracy and access requirements, in contrast with the simplification desired by the legislator?
Roberto Corbo: This is one of the great dilemmas of our country: too often we end up creating rules which, instead of helping us, risk further complicating the work of companies. Today we need transparent procedures and effective controls, but at the same time we must guarantee maximum speed in decisions and execution of interventions. Enhancing the solidity of a company does not have to mean introducing new bureaucratic obstacles.
Artificial intelligence, moreover, can be a valid support in the management of procedures, planning and budgeting. We must therefore use technology to simplify and accelerate, while keeping people’s responsibility and competence at the centre. Speed is fundamental, but it cannot be to the detriment of the quality and reliability of those who carry out the works.
