The project
On 19 August we published our critique of the Sisto reform of workplace safety. Those who read it will find a long, textual analysis, conducted provision by provision on the draft articles of the ministerial Commission. But that critique is not an isolated article: it is the second pillar of a project which, with today’s publication, is officially born, and which we here declare in full — the critique of corporate criminal law.
The first pillar we have already built: the reform of the liability of legal entities, Decree 231, rewritten by the Fidelbo Committee. The second came out on 19 August: the reform of workplace safety, rewritten by the Sisto Commission — two commissions established by the same Ministry six weeks apart, with one member in common and no reference to each other. The third is already under construction: the Italian law on artificial intelligence, Law 132 of 2025, with its symbolic criminal provisions written never to be applied. The fourth has just been served up by the legislator: the transposition of the European directive on the criminal protection of the environment, Legislative Decree 81 of 2026 — the same number, by an irony of fate, as the consolidated workplace-safety act —, which answered Europe’s request to fine companies up to five per cent of worldwide turnover by keeping its penalty quotas: at most, not even two million euros. And in due course the rest will come: the markets, taxation, competition — wherever criminal law meets the corporation.
Why a project, and not a series of articles? Because the building sites, looked at one at a time, let themselves be told as technicality: a threshold of punishability here, a ground of extinction there, a mitigating circumstance, a certification. It is their way of passing unnoticed. Looked at together, they show what each one, on its own, manages to hide: a direction. The same hand, the same method, the same beneficiary. We will look at them together.
The distortion: from the criminal law of labour to the criminal law of the employer
The criminal law of labour was born for one reason only: to protect, with the harshest weapon in the legal order, the life and health of those who work — that is, of those who, in the relations of production, endure the risk and do not govern it. The Constitution says that criminal liability is personal and that health is a fundamental right; the consolidated act says that the employer is the first guarantor of safety; the European framework directive makes the employer responsible “in all aspects connected with work”. This was — this is, on paper — the criminal law of labour: the genitive says who is protected.
The legislative season now under way reverses the genitive. The Sisto project, as we wrote in the very title of the critique published on 19 August, does not protect labour: it protects the employer. And it protects the employer — the formula is the most exact we have found, and we adopt it as the insignia of this section of the analysis — from the criminal law that ought to protect workers’ lives. The instrument no longer protects the working person from business risk: it is the business that is protected from the instrument. The protection remains criminal; what changes is who is protected. From the protection, through criminal law, of labour — to the protection, still through criminal law, of the employer from trial.
The mechanism, across the various building sites, is always the same and always presented as modernity: nothing is repealed, nothing is decriminalised — that would be visible, and would carry a political price. The offence is left where it is, indeed its penalties are raised for the shop window; and upstream, the device is built that prevents the trial from reaching the top. In the reformed Decree 231, the rewarding exits and the extinction of the offence for the entity that has reorganised itself. In the Sisto project, the gross-negligence shield for the employer who has adopted the model. In the artificial-intelligence law, offences drafted to catch no one. In the environment, penalties held at a level which, for a multinational, is a line in the accounts. Corporate criminal law remains standing as stage scenery; what empties is the stalls of the punishable — of those, that is, who sit at the top.
The second tendency: the securitisation of the duty of prevention
There is a word from finance that describes with precision what is happening to the duty of prevention, and we adopt it as a category of this project: securitisation. Finance gives that name to the operation that transforms a claim — a real relationship, made of a flesh-and-blood debtor and his ability to pay — into a paper security that circulates on its own, is bought and sold, and is worth not what it contains but the rating that someone, paid by the issuer, has stamped upon it. How that ended, when the world realised that the paper was worth more than the claims it was supposed to contain, we all remember.
The corporate criminal legislator is doing exactly this to the duty of prevention. The employer’s duty — to assess the risk, to remove it, to maintain the machines, to train the people, to spend what it takes — was a real relationship: an obligation to act, measured on the facts, sanctioned by the courts. The season under way transforms it, piece by piece, into paper: the organisational model, the risk-assessment document, the attestation of the joint body, the certification of the accredited body. Then it makes the paper a security: the model conforming to the standard in the shop window is presumed adequate by law. Then it gives the security a rating: the certifier’s stamp — chosen and paid, like the rating agencies of that era, by the issuer of the paper. And finally it grants the security the power that no paper should ever have: to extinguish liability. Once the paper is bought, the entity is exempt; once the paper is bought, the employer answers only for a gross negligence drawn on a nearly empty set. Pay for one, get two.
Real prevention — the money spent on maintenance, the plants shut down, the shifts lightened, the equipment handed out — is no longer the content of the duty: it is, at most, the forgotten underlying of a security that circulates on its own. And like every securitisation, this one too has its default already written. Except that here the default is not counted in basis points: it is counted in bodies. The risk, as we wrote in the closing lines of that article, is the one thing that does not shift and cannot be securitised — it stays in the air of the factory, in the water of the aquifer, in the blood of those who work.
And at the top of the paper chain stands the third element of the design, the oldest one: impunity at the top. Liability, extinguished at the summit by the model, descends the organisation chart, delegation by delegation, until it finds the accusable — the manager, the supervisor, the technician who signed the last piece of paper without holding the first euro. In that article we said it in the Latin that tradition demands, and the Latin remains the insignia of the whole project: absolvitur societas, absolvitur dominus — manet caper emissarius. The company absolved, the master absolved: the scapegoat remains.
The critiques of the Sisto reform, in twelve points
Those who want the full analysis will find it in the article published on 19 August, with the sources line by line. Here we state the critiques, all of them, in the order in which the text builds them.
First. The reform is born without numbers. In 2025 the reported fatal accidents at work were 1,093 — three deaths a day, a flat line since 2019 —; reported occupational diseases exploded from 45,000 to 98,463 in five years, up 119 per cent, the highest since the Seventies. And beneath the recorded figures lies the submerged part that science sees and the State does not: an estimated 17,000 occupational cancers a year against fewer than 940 recognised; 4,400 asbestos deaths estimated by the Istituto Superiore di Sanità, the national public-health institute (ISS), against 1,160 recognised; PFAS — a known human carcinogen, ubiquitous, eternal — which in the Italian statistics of occupational disease simply do not exist: one single official death, Pasqualino Zenere, recognised in 2025, while the University of Padua study with the ISS estimates almost four thousand excess deaths in the Veneto red zone alone. Of all this, in the Commission’s report, there is no trace: the problem the reform claims to solve is never measured.
Second. The Commission is built without the side of the dead. Twelve members: criminal lawyers, technicians, the public insurer, business — Confindustria twice, as a member and as a body heard in the consultations. No workers’ representative, no victims’ association, no inspector, no labour judge. And a single empirical basis: the Accredia study — produced by the body that accredits the certifiers of the very systems the study evaluates. A correlation, not a cause; a single source, and not an independent one.
Third. The heart of the reform is a safe-conduct. The new Art. 590-septies: if the entity has adopted an “adequate” model, the employer answers for the death and injuries of his workers only in cases of gross negligence. Ordinary negligence — the carelessness, imprudence, lack of skill that caused the death — becomes criminally irrelevant. What gross negligence is, the text does not say: it lists it by examples, and the examples are total omissions — having done nothing at all, the idiot-employer who does not occur in nature. A pneumatic void. Everything else — the assessment done badly, the maintenance skimped, the oversight loosened — sits below the threshold, with the purchased model there to testify for the defence. If the text stays as it is, the appointment is before the Constitutional Court.
Fourth. Impunity can be bought in a shop, and the shop window was dressed by the legislator. Art. 30, paragraph 5, of the consolidated act presumes conformity for the model drafted to the standard in the window — yesterday the UNI-INAIL guidelines and the British Standard, tomorrow ISO 45001. The chain has four links, each written in a different provision: the presumption of conformity; the entity’s exemption; the employer’s shield; the gross-negligence valve, closed to whoever has bought the model. Weld the links together and the provision the project does not have the courage to write in a single article reads: model conforming to X, employer and company not punishable. Each link, on its own, is a rebuttable presumption; the whole chain, in practice, works as a presumption iuris et de iure. And the party who attests conformity is chosen and paid by the employer: whoever sells the model sells, in the same package, the proof of its goodness. We foresee throngs of certifiers, and a market at least as large as that of Decree 231.
Fifth. The scapegoat is the final product. The shield covers the employer alone; the model covers the entity; managers and supervisors remain punishable for ordinary negligence, with increased penalties; the head of the prevention service acquires offences of his own. As you descend the steps of corporate power, the criminally sanctioned duties climb: no one at the top, everyone below. The cascade of delegations will descend the organisation chart until it finds the accusable — and the scapegoat is not a manufacturing defect of the device: it is its final product, here renamed the Delegate.
Sixth. Decree 231 completes its historic slide. In 2001 it was born to break the dogma — societas delinquere non potest — and to add a responsible party beside the human being, not to replace him: two liabilities, so that the organisation would no longer serve as a screen. The Sisto project makes it come full circle: the entity’s model screens the human being and continues to exempt the entity. The instrument created against the screen becomes a screen twice over. And Art. 27 of the Constitution — “criminal liability is personal”, written as a guarantee: no one punished for another’s act — now lives as a privilege: the employer not punished for his own act, provided the entity has adopted the right paper.
Seventh. The risk assessment is securitised. The employer answers if the document does not exist; if it exists and is mere paper, he no longer answers for it: the new Art. 31 declares that the employer “does not answer for the inadequacy of the risk assessment” if he has fulfilled the organisational obligations, and the new Art. 58-bis hands the head of the prevention service — the in-house technician, insurable, without one euro of spending power — offences of his own. The technical professions understood it before anyone else: they speak of a lightning conductor and of a hobbled guarantor.
Eighth. The State watches. The annual probability that an Italian firm receives a health-and-safety inspection is 1.1 per cent: once every eighty-nine years. Where inspections do happen, the irregularity rate has stood between 70 and 74 per cent for seven years. The “validated good practices”, to which the new indices of negligence refer, have been frozen since 27 November 2013. The reform adds not one inspector, not one filing duty, not one ex ante check: it perfects self-assessment paid for by the party assessed — privatisation with a hold-harmless clause. It is the class of phenomena that American scholarship calls State-facilitated crimes; with the Italian aggravating feature of a State that does not supervise, does not answer, and then joins the proceedings as a civil party against the accused company. For show, naturally.
Ninth. Europe demands, Italy exempts. Framework Directive 89/391 makes the employer responsible for safety “in all aspects connected with work”: a law that releases him for the mere paper adequacy of the model has, with that directive, a compatibility problem that will end up in Luxembourg. And on corporate penalties the calendar speaks for itself: Directive 2024/1203 requires maximum fines of no less than 3-5 per cent of worldwide turnover, or 24-40 million; the Italian transposition decree — in force since 2 June 2026 — rejects turnover “for proportionality” and keeps the quotas: for environmental disaster, not even two million. Three weeks later, the Sisto report reaches the Minister with the exemption as its dowry. With one hand the directive is transposed downwards; with the other, exemption is granted.
Tenth. The coin has two faces. The symbolic one — the statutory penalties that rise — is for the public; the effective one — the stalls of the punishable emptying — is for those who have equipped themselves. And beside the increased penalties for the delegates, a new mitigating circumstance for the event that “is not the exclusive consequence” of the conduct: as if an event ever could be, as if Art. 41 of the Code and two centuries of the equivalence of conditions had never been written. It is the botched answer to the most serious problem the reform crosses — the multifactorial nature of occupational disease —, a problem that belongs to the domain of risk and not to that of the necessary condition, and that demanded the prudent, considered passage to the Risk Society: not a discount on punishment, but new science — the epidemiology that measures risks — and new doctrine. The backwardness of our criminal-law culture, here, is palpable.
Eleventh. The tragedy under way has not reached the table. While occupational disease explodes, the word “exposure” does not appear in the text, nor the word “carcinogen”, nor any provision on health surveillance. Eighty kilometres from the ministerial tables, the PFAS case was meanwhile taking the legislator’s measure: the guinea-pig workers of Trissino and Spinetta Marengo with the highest PFOA blood burdens ever documented; the multinationals absent from the Vicenza trial — never charged, not acquitted: never judged —; no occupational exposure limit, no collective instrument for the victims, the drinking-water limits postponed “to give operators time”. Before an estimated four thousand deaths, silence is not neutrality: it is taking a side.
Twelfth. The judgment. This project is not a reform of workplace safety; it is a staggering retreat of the criminal liability of the employer and the company for death and disease at work. Risk is assessed as before, by the firm alone; it is checked as before, almost never; what changes is what happens after the death. We save the least bad page — the fast-track “red code” extended to workplace offences, the victim who enters the proceedings — and we record it as the exception that confirms: it is politically the most fragile, because it has not one organised defender; the shield has them all. And if the Sisto project were to become law together with its twin on Decree 231, the Italian legal order would have built, for death at work, the most complete exemption in Europe — and called it prevention.
What is to come
The project continues. The next chapter is already written and concerns artificial intelligence: Law 132 of 2025, its criminal law of paper, the paper gold of its promises. Then the environment, where the downward transposition of Directive 2024/1203 deserves a trial of its own. Then, in due course, the rest of corporate criminal law — wherever we find the same design, we will call it by its name.
The method remains what it has always been, and we declare it as a pact with our readers: official texts read word by word, numbers with their source line by line, derivations declared as such, the register of absence — the search conducted, and the negative result placed on the record — and every reported error verified and, where founded, corrected with an account given.
Because one thing, in all this paper, does not shift and cannot be securitised: the risk. It stays where it has always been — in the body of those who work. The guinea pigs do not shift. They are still waiting for someone to count them; and if the State does not defend the elementary goods of the human person, what is it for? We at whitecollarcrimes.it will see this passage through to the end.
whitecollarcrimes
a collective byline · guarantor: Avv. Luca Santa Maria · 28 August 2026