Artificial Intelligence and Business Crisis Prevention: What Changes with the New Code

Risk prevention is the fulcrum around which the new Corporate Crisis Code revolves. IT and AI systems play a central role in signaling the signs of a state of difficulty and allowing timely intervention aimed at resuming ordinary economic activity. That's how.

The new code of the crisis: the inspiring criteria of the reform

With the entry into force of the Code of business crisis and insolvency: alongside the provisions amending the Civil Code, already in force since March 2019, the entire body of legislation dedicated to the regulation of procedures has finally entered the legal system insolvency, intended to replace the provisions of the Bankruptcy Law [2] and to bring together under the same roof the other provisions on insolvency matters contained in special laws [3] .

As has been highlighted by many, the inspiring criteria of the reform consisted, on the one hand, of an attempt to overcome, even on the lexical level, the social stigma that over time has characterized crisis situations, which are now seen as a possible - albeit undesirable - element of entrepreneurial life [4]; on the other hand, it has consistently made residual the solution of the compulsory liquidation of assets aimed at satisfying creditors (the "old" bankruptcy), giving the entrepreneur the possibility of overcoming the state of crisis through the use of other instruments governed by the Code ( including, in particular, the negotiated settlement of the crisis, the restructuring agreements and the arrangement with creditors). Tools whose application, moreover, presupposes the ability to grasp the signs of crisis in advance, allowing to intervene when the same is reversible or, at least, controllable by the entrepreneur.

Appropriate structures in an anti-crisis function; the notion of crisis

Precisely in order to allow timely intervention at the first signs of difficulty , the Crisis Code has placed particular emphasis on the need for the entrepreneur to have adequate organizational, administrative and accounting structures to prevent the crisis.

On the one hand, in fact, it introduced a second paragraph to art. 2086 of the Italian Civil Code (already in force since March 16, 2019 ) , by virtue of which it has provided, for all companies organized in corporate or collective form: the crisis and the loss of business continuity, as well as ii) the duty to act without delay, using one of the tools provided for by the Crisis Code, in order to overcome the state of crisis and recover business continuity.

In other words, the law postulates that companies and enterprises organized collectively are equipped with organizational structures and administrative and accounting safeguards such as to allow not only compliance with the law and the articles of association, but also efficient and effective management [5 ] . Moreover, this is not the place to investigate how this obligation must be declined, also in consideration of the fact that, on the one hand, the law is completely silent on the content of the assets (except for what will be said below regarding the anti-crisis); and, on the other hand, that the structures must be tailor-made for each individual company to which they apply [6] .

The standard finds its completion today in art. 3, Crisis Code, which - in reiterating the obligation of collective entrepreneurs to set up structures - also establishes that the individual entrepreneur must also adopt suitable measures to promptly detect the state of crisis and immediately take the appropriate initiatives to make it forehead [7] .

What is meant by "crisis" is art. 2, paragraph 1, lett. a., Crisis Code to clarify it: the rule, following the changes made by Legislative Decree n. 83/2022, establishes that by "crisis" we mean the state of the debtor that makes insolvency possible - that is, the inability to regularly meet their obligations (see Article 2, paragraph 1, letter b., Of the Italian Civil Code crisis) - and which manifests itself in the inability to meet prospective cash flows to meet their obligations over the next twelve months.

Artificial intelligence and crisis prevention

In the light of the foregoing, it is quite clear that the structures must allow the entrepreneur to intercept the risk of crisis in advance . To this end, art. 3, Crisis Code, paragraphs 3 and 4 (introduced by Legislative Decree no. 83/2022) orientates the content of the assets on a finalistic level and lists some indices that the assets must be able to intercept in order to promptly foresee the emergence of the state of crisis. In particular, the third paragraph provides that the assets must be able to:

  1. detect imbalances of a patrimonial or economic-financial nature, in relation to the characteristics of the company;
  2. verify the sustainability of the debts in the following twelve months and intercept any signs of crisis indicated by the fourth paragraph of the regulation;
  3. make it possible to obtain the data necessary for carrying out the practical test for verifying the reasonable prosecution of the recovery in the context of the negotiated settlement procedure of the crisis.

The fourth paragraph, on the other hand, includes the existence of:

  • payables overdue for more than thirty days, for an amount greater than half of the total amount due by way of remuneration;
  • trade payables overdue for more than ninety days, for an amount greater than those not overdue;
  • exposures to banks or financial intermediaries overdue by more than sixty days or which by the same term have exceeded the limit of the credit lines obtained (and provided that they represent at least 5% of the total exposures);
  • social security and tax payables , to the extent provided for by art. 25 novies , paragraph 1, Crisis Code [8] .

From the combined provisions of Articles 2 and 3 of the Crisis Code, it emerges, on the one hand, how the notion of "crisis" is anchored to objective parameters , in the detection of which a decisive role can be played by computer and artificial intelligence systems.

This role can be played on several levels.

On a more elementary level - so to speak - in the preparation of the structures, care should be taken to use suitable IT tools both to collect and aggregate forecast data on the expected financial flows in relation to the obligations assumed, and to monitor the overall debt exposure. of the company, in order to have a concise but clear picture of the performance of the business and the possible existence of symptoms anticipating the crisis.

At a higher level, then, it could be envisaged - in the context of structures with a more refined architecture - the establishment of artificial intelligence systems capable of autonomously reprocessing the acquired data (as well as any additional data that the individual company considered relevant for check the "state of health" of one's business), in order not only to intercept the anticipatory signals of the crisis at the moment of their manifestation, but also - to a certain extent - to predict their occurrence.

Moreover, in the implementation of these systems, in order for them to be truly effective in preventing the crisis, care should be taken that they independently transmit the data, duly organized and systematized, to the management and control bodies [9] , in order to guarantee the best possible the administrative and supervisory functions entrusted to them. This is even more advisable when we consider that art. 25 octies, Crisis Code (in conjunction with Articles 12 and 17, Crisis Code), places on the control body the specific duty to report to the administrative body the need to access the measure of the negotiated settlement of the crisis in the the moment in which the existence of a situation of patrimonial or economic-financial imbalance that makes the crisis or insolvency probable and the recovery of the company can be reasonably pursued.

In order for this task to be carried out in the best possible way, therefore, it seems unavoidable that - within the anti-crisis structures - artificial intelligence systems are envisaged which, on the basis of the criteria set by art. 2 and 3, paragraphs 3 and 4 of the Crisis Code (briefly examined above), are able to signal the signs of the crisis in advance , in order to allow a timely and suitable intervention to allow the overcoming of the state of difficulty, as well as the recovery of ordinary economic activity.

With the caveat, however, that the administrative and control bodies will not be able to consider the respective obligations of information and management control fulfilled on the simple basis of the information provided by the IT systems, for at least two good reasons. The first is that computer processing depends - at least in an initial phase - on the quantity and quality of the information to be processed , so that, in the event of its insufficiency, the data produced by the machine would also have poor reliability. The second, relating above all to programs equipped with autonomous and unsupervised machine learning systems, concerns the possibility that the data processed autonomously by the softwarerisk being poorly intelligible [10] .

Therefore, in any case, it is not possible to think of a sort of de-responsibility of the corporate bodies for having adapted their behavior and their decisions to the results of the information system: a complete verification of the data will always be necessary to establish whether, indeed, from them the prodromal signs of the crisis can be found or not. As well as directors and statutory auditors, they will also have to assess whether, in spite of any positive computer indications, the company risks finding itself in crisis and therefore requires the activation of the insolvency procedures provided for by the Code.

The practical test for verifying the reasonable prosecution of the recovery

We have seen, in the previous lines, that the structures must be able to collect the information necessary for carrying out the practical test relating to the prosecution of the recovery in the hypothesis of access to the negotiated settlement procedure of the business crisis.

And precisely this test constitutes - in hindsight - the first, embryonic, starting point for the computerization of the crisis contained in the text of the Code of the crisis [11] . In fact, this test, which is made available on the electronic platform dedicated to negotiated settlement procedures, consists in the computerized processing of the relationship between certain categories of overdue payables - precisely indicated by the executive decree of 28.9.2021 of the Ministry of Justice - and the annual flow to service these payables. In the intent of the legislator, a lower test result corresponds to a greater probability of recovery of the company, through negotiations with creditors or extraordinary initiatives; on the other hand, a high ratio between debts and active flows is symptomatic of a state of crisis that can only be overcome indirectly, through the sale or rental of the company.

Now, it is quite clear that, in order to make the strategies for overcoming the crisis more effective, the entrepreneur must be able, even before accessing the telematic platform for negotiated settlement, to independently process such data, in order to check which initiative to take.

In this perspective, it is well understood why the legislator requires the structures to be able to collect all the data necessary to support the practical test on reasonable recovery: they, in fact, are suitably "filtered" by the sensitivity and competence of the administrative body , they are helpful both in verifying the existence of the state of crisis, and in identifying the most suitable tool to overcome it. And, indeed, precisely in the phase of choosing the most suitable insolvency procedure, further space could be carved out for artificial intelligence which, on the basis of company data, could suggest the most suitable solution to the crisis. An instrument that could be chosen - but these are suggestions still to be explored - even in the same way as any predictive justice algorithms which, in the case, allow to evaluate the "stability" of the chosen solution even in court.

  1. For an overview of the interventions that have taken place on the original text of the Crisis Code, as well as the extensions to the entry into force of the same, see Garesio, Brief notes on the state of the art of bankruptcy law , in , 2022, following which Legislative Decree no. 83/2022 which made the latest changes to the Crisis Code, definitively putting it into effect on 15.7.2022. 
  2. The art. 389 of the Crisis Code provides that for all the procedures pending before the entry into force of the Crisis Code, the rules envisaged by the Bankruptcy Law and by the additional laws on insolvency matters applicable at the time of the initiation of the relative procedure are fully valid. The result is a coexistence of disciplines, between the Bankruptcy Law and the Crisis Code, which could last for a few years, generating not a few uncertainties for operators. 
  3. We refer, in particular, to the regulation of over-indebtedness, already contained in Law no. 3/2012 and today governed by arts. 65 and following and 268 ss., Crisis Code, as well as the rules on the negotiated settlement of the business crisis, originally envisaged in Legislative Decree no. 118/2021 and now substantially transposed into Title II of Part One, Crisis Code (replacing the assisted settlement procedure of the crisis, initially regulated therein ). 
  4. This is exemplified, even only on the definitory level, by the fact that the bankruptcy procedure is now called "judicial liquidation" 
  5. See Meruzzi, The adequacy of the structures , in Adequate structures and organizational models in the corporate governance of joint-stock companies , edited by Irrera, Bologna, 2016, 43 ss. 
  6. Nonetheless, for a more complete analysis of the obligation of structures and their discipline, as well as for further references and citations, see for all Irrera, Adequacy of the organizational, administrative and accounting structure , in Aa. Vv., Treaty of companies , edited by Donativi, in course of publication for the types of Utet. 
  7. Which, in fact, also entails the obligation for the individual entrepreneur to equip himself with structures in order to combat the crisis. In this sense v. for all Cagnasso, The suitable measures, the adequate structures and the organization of business activity , in Nuovo dir. company , 2021, 1597 ff. 
  8. Which, following the changes made by Law 4.8.2022, n. 122 (converting, with amendments, of Legislative Decree no. 73/2022), provides that: « [1]. The National Institute of Social Security, the National Institute for Insurance against Accidents at Work, the Revenue Agency and the Revenue-Collection Agency report to the entrepreneur and, where applicable, to the supervisory body, in the person of the chairman of the board of statutory auditors in the case of a collegial body, by certified e-mail or, failing that, by registered letter with acknowledgment of receipt sent to the address resulting from the tax register:
    1. a) for the National Institute of Social Security, the delay of more than ninety days in the payment of social security contributions of a higher amount:

    1) for companies with subordinate and para-subordinate workers, to 30 percent of those due in the previous year and to the amount of € 15,000;

    2) for companies without subordinate and para-subordinate workers, to the amount of 5,000 euros;

    1. b) for the National Institute for Insurance against Accidents at Work, the existence of a debt for insurance premiums overdue for more than ninety days and not paid exceeding the amount of 5,000 euros;
    2. c) for the Revenue Agency, the existence of an overdue and unpaid debt relating to the value added tax, resulting from the communication of the data of the periodic settlements referred to in article 21-bis of the decree-law of 31 May 2010 , no. 78, converted, with modifications, by law 30 July 2010, n. 122, higher than the amount of 5,000 euros and, in any case, not less than 10 percent of the amount of the turnover resulting from the return relating to the previous tax year; the report is in any case sent if the debt is greater than the amount of € 20,000;

    d) for the Revenue-Collection Agency, the existence of credits entrusted for collection, self-declared or definitively ascertained and expired for more than ninety days, higher, for sole proprietorships, than the amount of 100,000 euros, for companies of persons, to the amount of 200,000 euros and, for other companies, to the amount of 500,000 euros  

  9. On the relationship between artificial intelligence and the information obligations of the corporate bodies, see the interesting essay by Scarabelli, Artificial intelligence and information flows within the board of directors , in Nuovo dir. company , 2022, 599 ff. 
  10. For these risks, as well as for a wider exposure of the critical profiles linked to the automation of information flows, the reference is once again to Scarabelli, Op. Cit. , 616 ff. 
  11. And, before then, in Legislative Decree no. 118/2021, which introduced the negotiated settlement procedure which today merged into the Crisis Code. 

Do you want more information?

Article Rating: 
Average: 4.9 (8 votes)
Total reads: 1,119
Marco Catalano's picture

Collaborates with the Chair of Civil Procedure at the University of Turin, and has written articles, legal case notes and commentaries, predominantly regarding the sector of Arbitration Law.

He is a member of the editing staff of “NDS - Il Nuovo Diritto delle Società”.

Areas of Practice

Economic Sectors