A philosopher pins corporate negligence on the roles that exist to check the facts

The route from a company's false position to a named person's negligence runs through job descriptions. Ludwig sorts roles into responsibility magnets and responsibility shelters. A safety inspector, or a board overseeing a CEO, exists specifically to catch bad information before it drives a decision, and carries a duty to determine facts or check assumptions ^[18].

"Corporations and other organizations are often said to want, believe, and intend various things," Ludwig writes in the study ^[23]. On his analysis that talk gives information about what an organization will do in circumstances relevant to corporate interests, grounded in the commitments its employees have made in their official roles ^[13]. Volkswagen has said publicly that it believes quantum computing could transform how the company uses data ^[14]. Ludwig's reading of a statement like that is narrow: people in roles with authority to commit company resources have funded research, assigned engineers and briefed investors, or are poised to. The rest of the workforce need not agree ^[15].

"Talk of corporate beliefs and other attitudes is a way of capturing that pattern of commitments," Ludwig said ^[16]. "It is the appropriation of a vocabulary we apply to individuals for a different purpose" ^[17].

The film set gives him the cleanest illustration. On a set with a clear division of labor, the duty to verify the ammunition belonged to the armorer and not to everyone present. That formal division of labor creates a matching division of moral responsibility ^[20]. Ludwig applies the principle to the more complex formal division of labor inside a corporation ^[21]. Takata introduced a new, smaller but catastrophically flawed airbag design, and on his account those flaws should have been detected by the people charged with design and testing ^[22]. The inflators the company kept clearing exploded on impact and triggered the largest automotive recall in U.S. history ^[2].

The support for all of this is three real harm cases and one hypothetical committee ^[1]. Conceptual analysis normally rests on cases like these. The fair test is whether the proposal meets constraints its rivals fail, including Ludwig's own condition that the analysis must not end in the unwanted result that no individual is ever to blame in cases of corporate negligence ^[7]. It meets them, though the paper does not settle whether the magnet in a particular company can be identified from the documents an investigator or a jury will actually see.