Publications

(ITA) Il contributo suggerisce l’affermarsi di un moderno principio di residualità non limitato ad informare il rapporto tra regolazione e concorrenza, ma che richiede una giustificazione dell’intervento pubblico in termini di effettiva necessità e ragioni specifiche che ne sono alla base, per spingersi ad informare contenuto regolatorio. Le ragioni risultano non solo arricchite rispetto a quelle tradizionali, ma anche disancorate da problemi già concretizzatisi per abbracciare una visione prospettica e anticipatoria. --- (ENG) The paper suggests the rise of a modern principle of residuality, which exceeds the balance between regulation and competition. This principle calls for the justification of regulation in terms of necessity. Problem drivers should be also taken into consideration, as well as the regulatory content. The problem drivers appear to be enriched compared to the traditional ones, whilst also being disentangled from problems that have already materialized so as to embrace a prospective and anticipatory approach.

These two volumes collect twenty five articles and papers published within the “Governance of/through Data” research project financed by the Italian Ministry of Universities. The research project, which was promoted by Roma Tre University, as project lead, and saw the participation of professors and reseachers from Bocconi University in Milan; LUMSA University in Rome; Salento University in Lecce and Turin Polytechnic, cover multiple issues which are here presented in five sections: Algorithms and artificial intelligence; Antitrust, artificial intelligence and data; Big Data; Data governance; Data protection and privacy.

The AI Index is an independent initiative at the Stanford Institute for Human-Centered Artificial Intelligence (HAI), led by the AI Index Steering Committee, an interdisciplinary group of experts from across academia and industry. The annual report tracks, collates, distills, and visualizes data relating to artificial intelligence, enabling decision-makers to take meaningful action to advance AI responsibly and ethically with humans in mind. The AI Index collaborates with many different organizations to track progress in artificial intelligence. These organizations include: the Center for Security and Emerging Technology at Georgetown University, LinkedIn, NetBase Quid, Lightcast, and McKinsey. The 2023 report also features more self-collected data and original analysis than ever before. This year’s report included new analysis on foundation models, including their geopolitics and training costs, the environmental impact of AI systems, K-12 AI education, and public opinion trends in AI. The AI Index also broadened its tracking of global AI legislation from 25 countries in 2022 to 127 in 2023.

​

The “Gatekeeper” approach to competition policy proceeds by identifying a few large firms as gatekeepers. It then applies aggressive competition rules to them while leaving others unaffected. Legislation that was considered last year in Congress would have done this. While that legislation failed to pass, the issue of competitive control of large firms remains alive and will certainly return. The proposed American Innovation and Choice Online Act illustrates the problems of gatekeeper approaches. First, it selects for harsh treatment a portion of the economy that is generally outperforming the rest. Second, it limits its domain to online firms without any basis for thinking that competition problems are more serious in that portion of the market. Third, by defining covered firms by large firm size rather than product market share it misses anticompetitive actions by firms that are smaller overall but dominant in particular products. Finally, for those firms identified as gatekeepers it would prohibit a great deal of competitively harmless activity.

Antitrust law addresses conspiracy, or collaborative conduct, more harshly than it does unilateral conduct. One person acting alone can get away with far more than groups of firms acting by agreement. In most cases that distinction is justified. Creating substantial market power unilaterally is difficult and relatively uncommon, but it can be created in a moment’s time by an agreement among firms. But how do antitrust tribunals determine when conduct is unilateral rather than collaborative? Often the ansawer is obvious, but sometimes it is not. Two statutory provisions were intended to be the umpire of such decisions. A section of the Sherman considered so important that it was re-enacted in the Clayton Act provides that corporations and associations authorized by state law should be treated as “persons,” or single actors. The provisions address the core problems about internal corporate structure, including the single-entity status of holding companies, the legitimacy or not of suits between shareholders or employees and their firm, or the status of professional associations. The fact that the Sherman Act’s corporate personhood provision was re-enacted virtually verbatim in the Clayton Act is significant, because the intervening quarter century had witnessed a fierce debate over the power and reach of the business corporation. The personhood provisions fall short, however, because they completely ignore most of the interesting cases where conspiratorial capacity is in issue. They have nothing to contribute to situations where the precise boundaries of the corporation become ambiguous. Nor do they provide a solution to the problem of how to address labor disputes between an employer and its own employees. Further, and inadvertently, the statutes have encouraged certain types of industry structures that are not mandated by good competition policy, including the tendency to merge in order to avoid harsh rules about collusion, and the tendency to integrate vertically by ownership even when contractual integration might be superior.

Machine learning, or artificial intelligence, refers to a vast array of different algorithms that are being put to highly varied uses, including in transportation, medicine, social media, marketing, and many other settings. Not only do machine-learning algorithms vary widely across their types and uses, but they are evolving constantly. Even the same algorithm can perform quite differently over time as it is fed new data. Due to the staggering heterogeneity of these algorithms, multiple regulatory agencies will be needed to regulate the use of machine learning, each within their own discrete area of specialization. Even these specialized expert agencies, though, will still face the challenge of heterogeneity and must approach their task of regulating machine learning with agility. They must build up their capacity in data sciences, deploy flexible strategies such as management-based regulation, and remain constantly vigilant. Regulators should also consider how they can use machine-learning tools themselves to enhance their ability to protect the public from the adverse effects of machine learning. Effective regulatory governance of machine learning should be possible, but it will depend on the constant pursuit of regulatory excellence.

In this report, the National Board of Trade Sweden highlights the challenges that digital innovation creates for technical regulation. The report questions whether artificial intelligence and cyber vulnerabilities are changing the way industrial goods, such as some smartphones, medical devices, and vehicles with embedded digital technologies, are regulated.

​

​