Bartek Chomanski

Suppose it were possible to improve people’s ability to identify fake political news and resist common manipulative techniques, without at the same time interfering with their free speech rights. Suppose such improvements were easy to implement at scale on the most important social media platforms. It seems that doing so would obviously be morally permissible and, arguably, morally required. A number of real-world anti-misinformation interventions (from fact-checks to accuracy nudges) developed by psychologists in recent years appear to have just these features. They are mostly unobtrusive and non-censorial, moderately effective at improving our ability to spot fake news and resist manipulation, and they positively impact news sharing behavior. It is also feasible to implement them at scale. In this paper, I argue that, even if the counterstrategies to the spread of misinformation are as successful as their advocates claim, they are nevertheless morally questionable in important respects. Specifically, pursuing (a subset of) such strategies wrongs social media users by reducing their welfare, and all the interventions considered in the paper are _prima facie_ disrespectful. These concerns suggest that the interventions’ all-things-considered justification hinges on whether their benefits outweigh these considerations. It is often maintained that they unquestionably do: the interventions help prevent serious harms to democratic institutions. However, significant empirical evidence undercuts this justification of the interventions. Overall, the paper’s arguments suggest that implementing effective anti-misinformation interventions implicates moral issues—such as harm and disrespect to the recipients—which seem important for our assessment of the merits of the strategy, but remain largely absent in the philosophical literature on the topic, focused, understandably, on the epistemically best responses to the problem.

One powerful objection to recent arguments for greater state control over digital expression rests on the assumption of behavioral symmetry (meaning, roughly, that we should not introduce stark differences in motivational profiles between private actors and state officials). In this paper, I argue that the adoption of behavioral symmetry severely constrains the space of feasible institutional solutions to the problem of excessive censorship, including such remedies as the introduction of constitutional protections for speech, as well as inter-jurisdictional and intra-jurisdictional competition—sometimes favored as effective restraints on the censorial state. Instead, I suggest that, once we grant behavioral symmetry, we are led toward the conclusion that an anti-censorship ethos—a complex of ideas that sustain the intellectual appeal of free expression—is a necessary condition for digital speech to be protected.

Libertarians and their fellow travellers exhibit an ambivalent attitude towards large digital technology companies that have come to dominate internet commerce over the last two decades (the so-called Big Tech). On the one hand, some praise their liberating potential and hail their success as proof that relatively unregulated markets can produce goods and services of enormous value to the consumer. Others, however, decry their censoriousness and an overly deferential attitude to governments’ wishes, sometimes going as far as to compare Big Tech’s enforcement of speech policies to Soviet-style gulag systems. Unfortunately, this internal – to libertarianism – squabble is of limited usefulness to the larger philosophical discussion about the ethics and politics of Big Tech, where appeals to strict private property rights or illegitimacy of government influence on corporations are likely to fall on deaf ears. Yet, concerns in the more mainstream literature on the ethics and governance of digital markets offer a serious challenge to a largely non-interventionist ethos of libertarianism. Libertarians, therefore, would do well to address them on their own terms instead. This is what I seek to begin to do in this chapter. I will start with a brief overview of approaches to digital technology that have emerged in the (very broadly construed) libertarian scholarship. I will then present and address the arguments that seem to me to be among the most pressing examples of social harms allegedly caused by the massive expansion of digital markets: harms to democratic institutions (including the contribution to political violence and extremism), harms to the enforcement of public health measures, and harms to children. I will end with a discussion of the argument that tech policy discourse is polluted in virtue of many researchers, conferences, and programs being funded by Big Tech companies. I will argue that these arguments rest either on weak empirical foundations or rely on implausible normative principles (or both), and that one need not be a libertarian to concede this.

Concerns about artificial intelligence (AI) and its potential existential risks have garnered significant attention, with figures like Geoffrey Hinton and Dennis Hassabis advocating for robust safeguards against catastrophic outcomes. Prominent scholars, such as Nick Bostrom and Max Tegmark, have further advanced the discourse by exploring the long-term impacts of superintelligent AI. However, this existential risk narrative faces criticism, particularly in popular media, where scholars like Timnit Gebru, Melanie Mitchell, and Nick Clegg argue, among other things, that it distracts from pressing current issues. Despite extensive media coverage, skepticism toward the existential risk discourse has received limited rigorous treatment in academic literature. Addressing this imbalance, this paper reconstructs and evaluates three common arguments against the existential risk perspective: the Distraction Argument, the Argument from Human Frailty, and the Checkpoints for Intervention Argument. By systematically reconstructing and assessing these arguments, the paper aims to provide a foundation for more balanced academic discourse and further research on AI.

This paper makes the following argument: (1) If there is political bias in implementing misinformation-countering measures, then the measures will asymmetrically target the cluster of misinformation associated with the disfavored side. (2) Given such asymmetric targeting, it is likely that the measures will induce significant trust disparity between the sides, and cause self-censorship and withdrawal by people supporting the disfavored side. (3) If there is trust disparity in an epistemic environment, then it is likely that the environment will degenerate into an echo chamber. (4) If there is self-censorship and withdrawal, then there will be a reduction in political expression. (5) It is likely that there will be bias in implementing misinformation-countering measures. (6) Therefore, it is likely that anti-misinformation interventions will introduce an echo-chamber-like epistemic dynamics in environments where the interventions are implemented, and that they will cause a reduction in political expression. The conclusion stands even if the interventions successfully achieve their goals, and should foster a much more serious reflection on the all-things-considered desirability of anti-misinformation measures.

In an interview for Playboy in 1969, media scientist Marshall McLuhan, known for, among other things, the expression that the medium is the message, claimed that computers will be able to orchestrate the lives of many people in the near future. Machines, the Canadian thinker said, will be able to take over the main media channels, write messages themselves, and distribute them to the population. Meanwhile, we know that McLuhan’s prediction was not complete nonsense—think of Microsoft’s chatbot TAY (the acronym stands for Thinking About You) that sent sexist and racist messages into the world via Twitter in March 2006. At the time, however, interviewer Eric Norden was somewhat impressed by what McLuhan was claiming. He asked if we as humanity will still have control over technology. Or, Norden queried, will technology control humanity? McLuhan replied as follows: “I see no possibility of a worldwide Luddite rebellion that will smash all machinery to bits, so we might as well sit back and see what is happening and what will happen to us in a cybernetic world. Resenting a new technology will not halt its progress.” (Norden, 1969).

The First Industrial Revolution relied on steam, the Second on electricity, and the Third on computers and the internet. For several years now, we have been in a new revolution. This has a lot to do with what the internet provided us with: data, big data, actually an awful lot of data about an awful lot of things from our online and offline lives. To give you an idea: it is said that 90% of the data available to us today was generated in the last 5 years. ‘Data is the new oil’ has been the cliché since 2006, when mathematician Clive Humby first used the phrase. And the USA is the new Saudi Arabia. What systems are being fueled with data?

This paper critically examines Giles Howdle’s (2023) view that microtargeting of political advertisements is morally problematic, regardless of the content of the ad. We offer three reasons to be skeptical of this conclusion: the wrongness of microtargeting is highly context-sensitive in a way that undermines Howdle’s strong thesis, the practice seems unlikely to affect voters in the way Howdle supposes, and the ads’ effectiveness in preventing deliberation is inversely proportional to how interesting and engaging their content is. We take these reasons to put significant pressure on Howdle’s argument in a way that renders his bold and novel claims unjustified by the considerations put in their favor.

This paper argues that if the critics of the currently dominant notice-and-consent model of governing digital data transactions are correct, then they should oppose political reforms of the model. The crux of the argument is as follows: the reasons the critics give for doubting the effectiveness of notice-and-consent in protecting user privacy (namely, ordinary users’ various cognitive deficiencies and the inherent inscrutability of the subject matter) are also reasons for doubting the effectiveness of protecting user privacy through democratic or regulatory means. Furthermore, insofar as attempts to improve the notice-and-consent model through more legislation or regulation would also involve more coercion than the status quo, they should be resisted on normative grounds. So, out of the bad options we have when it comes to protecting digital privacy, it seems – contrary to the majority position advanced in the literature – that we should stick with notice-and-consent.

A number of scholars and policy-makers have raised serious concerns about the impact of chatbots and generative artificial intelligence (AI) on the spread of political disinformation. An increasingly popular proposal to address this concern is to pass laws that, by requiring that artificially generated and artificially disseminated content be labeled as such, aim to ensure a degree of transparency in this rapidly transforming environment. This article argues that such laws are misguided, for two reasons. We first aim to show that legally requiring the disclosure of the automated nature of bot accounts and AI-generated content is unlikely to succeed in improving the quality of political discussion on social media. This is because information that an account spreading or creating political information is a bot or a language model is itself politically relevant information, and people reason very poorly about such information. Second, we aim to show that the main motivation for these laws – the threat of coordinated disinformation campaigns (automated or not) – appears overstated.

It is frequently argued that false and misleading claims, spread primarily on social media, are a serious problem in need of urgent response. Current strategies to address the problem – relying on fact-checks, source labeling, limits on the visibility of certain claims, and, ultimately, content removals – face two serious shortcomings: they are ineffective and biased. Consequently, it is reasonable to want to seek alternatives. This paper provides one: to address the problems with misinformation, social media platforms should abandon third-party fact-checks and rely instead on user-driven prediction markets. This solution is likely less biased and more effective than currently implemented alternatives and, therefore, constitutes a superior way of tackling misinformation.

This paper argues that the well-established fact of political irrationality imposes substantial constraints on how governments may combat the threat of political misinformation. Though attempts at regulating misinformation are becoming increasingly popular, both among policymakers and theorists, I intend to show that, for a wide range of anti-misinformation interventions (collectively termed “debunking” and “source labeling”), these attempts ought to be abandoned. My argument relies primarily on the fact that most people process politically-relevant information in biased and motivated ways. Since debunking or factual correction of politically relevant misinformation (as well as source labeling) themselves consist of providing politically-relevant information, they are also very likely to be processed in irrational ways. This makes it extremely difficult to effectively correct people’s political beliefs and political information processing. Since governments should not pursue policies likely to be futile, they should refrain from mandating such interventions. My conclusion is of relevance to considerable literature in digital ethics of misinformation. It shows that many celebrated works in the field ignore political irrationality and fail to consider its implications.

Digital privacy scholars tend to bemoan ordinary people’s limited knowledge of and lukewarm interest in what happens to their digital data. This general lack of interest and knowledge is often taken as a consideration in favor of legislation aiming to force internet companies into adopting more responsible data practices. While we remain silent on whether any new laws are called for, in this paper we wish to underline a neglected consequence of people’s ignorance of and apathy for digital privacy: their potential to encourage capture by industry interests. In particular, we argue that such laws may be at increased risk of capture because they are unlikely to be democratically responsive. We make this claim on a twofold basis: first, well-known theoretical mechanisms explaining how the absence of responsiveness leads to capture, identified in prior political science and political philosophy literature, yield the prediction that digital privacy legislation is likely to be unresponsive and thus captured; second, empirical data concerning the European Union’s digital privacy laws, with a special focus on the General Data Protection Regulation, appears to confirm these predictions: the bloc’s (world’s?) flagship privacy protection law seems more responsive to corporate than citizen interests.

This paper argues, against the prevailing view, that consent to privacy policies that regular internet users usually give is largely unproblematic from the moral point of view. To substantiate this claim, we rely on the idea of the right not to know (RNTK), as developed by bioethicists. Defenders of the RNTK in bioethical literature on informed consent claim that patients generally have the right to refuse medically relevant information. In this article we extend the application of the RNTK to online privacy. We then argue that if internet users can be thought of as exercising their RNTK before consenting to privacy policies, their consent ought to be considered free of the standard charges leveled against it by critics.

Many leading intellectuals, technologists, commentators, and ordinary people have in recent weeks become embroiled in a fiery debate (yet to hit the pages of scholarly journals) on the alleged need to press pause on the development of generative artificial intelligence (AI). Spurred by an open letter from the Future of Life Institute (FLI) calling for just such a pause, the debate occasioned, at lightning speed, a large number of responses from a variety of sources pursuing a variety of argumentative strategies. Not all of the respondents resist the calls for the pause. For example, the Distributed AI Research Institute (DAIRI) has issued a statement claiming that while the pause is indeed desirable, the FLI’s focus on predicted existential risks to the exclusion of present-day concerns is misguided. The discussion shows no signs of abating. I wish to raise an objection both to the FLI’s open letter, and to the statement by DAIRI (I will refer to both collectively as “letters” or “open letters”). While in the context of the issues brought up by the two Institutes, this objection is, to my knowledge, novel, I pretend to little originality in voicing it. Rather, I offer a simple (yet unappreciated) application of some insights from political philosophy and economics to show that the recommendations contained in both missives are severely underargued.

In “Accepting Moral Responsibility for the Actions of Autonomous Weapons Systems—a Moral Gambit” (2022), Mariarosaria Taddeo and Alexander Blanchard answer one of the most vexing issues in current ethics of technology: how to close the so-called “responsibility gap”? Their solution is to require that autonomous weapons systems (AWSs) may only be used if there is some human being who accepts the ex ante responsibility for those actions of the AWS that could not have been predicted or intended (in such cases, the human being takes what the authors call the “moral gambit”). The authors then propose several institutional safeguards to implement in order to ensure that the moral gambit is taken in a fair and just way. This paper explores this suggestion in the context of the institutional settings within which AWSs are most likely to be deployed. It raises some concerns as to the feasibility of Taddeo and Blanchard’s proposal, in light of the recent empirical work on the incentive structures likely to exist within militaries. It then presents a potential problem that may arise in case the accountability mechanisms are successfully implemented.

Is it wrong to distract? Is it wrong to direct others’ attention in ways they otherwise would not choose? If so, what are the grounds of this wrong – and, in expounding them, do we have to at once condemn large chunks of contemporary digital commerce (also known as the attention economy)? In what follows, I attempt to cast light on these questions. Specifically, I argue – following the pioneering work of Jasper Tran and Anuj Puri – that there is a right to attention, and that its existence underlies some of our claims regarding the wrongness of distractions. However, I depart from both these authors in two respects: first, I present a new way of deriving the right to attention, grounding it in the more fundamental right to mental integrity. Second, I remain agnostic on whether the contemporary business practices of capturing attention in exchange for a variety of digital products and services are plagued by routine violations of the right.

It might become possible to build artificial minds with the capacity for experience. This raises a plethora of ethical issues, explored, among others, in the context of whole brain emulations (WBE). In this paper, I will take up the problem of vulnerability – given, for various reasons, less attention in the literature – that the conscious emulations will likely exhibit. Specifically, I will examine the role that vulnerability plays in generating ethical issues that may arise when dealing with WBEs. I will argue that concerns about vulnerability are more matters of institutional design than individual ethics, both when it comes to creating humanlike brain emulations, and when animal-like emulations are concerned. Consequently, the article contains reflection on some institutional measures that can be taken to protect the sims' interests. It concludes that an institutional framework more likely to succeed in this task is competitive and poly-centric, rather than monopolistic and centralized.

With the advent of automated decision-making, governments have increasingly begun to rely on artificially intelligent algorithms to inform policy decisions across a range of domains of government interest and influence. The practice has not gone unnoticed among philosophers, worried about “algocracy”, and its ethical and political impacts. One of the chief issues of ethical and political significance raised by algocratic governance, so the argument goes, is the lack of transparency of algorithms. One of the best-known examples of philosophical analyses of algocracy is John Danaher’s “The threat of algocracy”, arguing that government by algorithm undermines political legitimacy. In this paper, I will treat Danaher’s argument as a springboard for raising additional questions about the connections between algocracy, comprehensibility, and legitimacy, especially in light of empirical results about what we can expect the voters and policymakers to know. The paper has the following structure: in Sect. 2, I introduce the basics of Danaher’s argument regarding algocracy. In Sect. 3 I argue that the algocratic threat to legitimacy has troubling implications for social justice. In Sect. 4, I argue that, nevertheless, there seem to be good reasons for governments to rely on algorithmic decision support systems. Lastly, I try to resolve the apparent tension between the findings of the two preceding Sections.

Pyrrho and colleagues (2022) argue that the loss of health privacy can damage democratic values by increasing social polarization, removing individual choice, and limiting self-determination. As a remedy, the authors propose a data-regulation regime that prohibits companies from using such data for discriminatory purposes. Our commentary addresses three issues. First, we point out an additional problematic dimension of excessive health privacy loss, namely, the potential racialization of groups and individuals that it may likely contribute to. Second, we note that, in our view, the authors’ argument for more regulation rests on an invidious comparison between the realistically described status quo and the idealized picture of the imagined regulatory regime that the authors briefly propose. Third, we argue that, despite existing regulations, both private and government actors frequently use private data in ways that lead to ethically problematic outcomes, especially when it comes to racialized communities.

Must opponents of creating conscious artificial agents embrace anti-natalism? Must anti-natalists be against the creation of conscious artificial agents? This article examines three attempts to argue against the creation of potentially conscious artificial intelligence (AI) in the context of these questions. The examination reveals that the argumentative strategy each author pursues commits them to the anti-natalist position with respect to procreation; that is to say, each author's argument, if applied consistently, should lead them to embrace the conclusion that procreation is, at best, morally problematic. However, the article also argues that anti-natalists can find the production of some possible artificially conscious AI permissible. Thus, the creation of potentially conscious AI could be accepted by both friends and foes of anti-natalism.

Having been involved in a slew of recent scandals, many of the world’s largest technology companies (“Big Tech,” “Digital Titans”) embarked on devising numerous codes of ethics, intended to promote improved standards in the conduct of their business. These efforts have attracted largely critical interdisciplinary academic attention. The critics have identified the voluntary character of the industry ethics codes as among the main obstacles to their efficacy. This is because individual industry leaders and employees, flawed human beings that they are, cannot be relied on voluntarily to conform with what justice demands, especially when faced with powerful incentives to pursue their own self-interest instead. Consequently, the critics have recommended a suite of laws and regulations to force the tech companies to better comply with the requirements of justice. At the same time, they have paid little attention to the possibility that individuals acting within the political context, e.g. as lawmakers and regulators, are also imperfect and need not be wholly compliant with what justice demands. This paper argues that such an omission is far from trivial. It creates a heavy argumentative burden on the part of the critics that they by and large fail to discharge. As a result, the case for Big Tech regulation that emerges from the recent literature has substantial lacunae and more work needs to be done before we can accept the critics’ calls for greater state involvement in the industry.

In this paper, I outline a proposal for assigning liability for autonomous machines modeled on the doctrine of respondeat superior. I argue that the machines’ users’ or designers’ liability should be determined by the manner in which the machines are created, which, in turn, should be responsive to considerations of the machines’ welfare interests. This approach has the twin virtues of promoting socially beneficial design of machines, and of taking their potential moral patiency seriously. I then argue for abandoning the retributive approach to machine crime in favor of prioritizing restitution. I argue that this shift better conforms to what justice demands when sophisticated artificial agents of uncertain moral status are concerned.

In this paper, I provide an account of the spatiality of olfactory experiences in terms of topological properties. I argue that thinking of olfactory experiences as making the subject aware of topological properties enables us to address popular objections against the spatiality of smells, and it makes sense of everyday spatial olfactory phenomenology better than its competitors. I argue for this latter claim on the basis of reflection on thought experiments familiar from the philosophical literature on olfaction, as well as on the basis of some empirical data about the localization of smells. I conclude by suggesting how the naïve‐topology framework could be applied in debates about the spatiality of other types of experiences.

It could become technologically possible to build artificial agents instantiating whatever properties are sufficient for personhood. It is also possible, if not likely, that such beings could be built for commercial purposes. This paper asks whether such commercialization can be handled in a way that is not morally reprehensible, and answers in the affirmative. There exists a morally acceptable institutional framework that could allow for building artificial persons for commercial gain. The paper first considers the minimal ethical requirements that any system of commercial production of such artificial persons would have to meet. It then shows that it is possible for these requirements to be met, and that doing so will make the commercial production of artificial persons permissible. Lastly, it briefly presents one potential blueprint for how such a framework could look like—inspired by the real-world model of compensating the training of athletes—and then addresses some objections to the view.

In a stimulating recent article for this journal (van Wynsberghe and Robbins in Sci Eng Ethics 25(3):719–735, 2019), Aimee van Wynsberghe and Scott Robbins mount a serious critique of a number of reasons advanced in favor of building artificial moral agents (AMAs). In light of their critique, vW&R make two recommendations: they advocate a moratorium on the commercialization of AMAs and suggest that the argumentative burden is now shifted onto the proponents of AMAs to come up with new reasons for building them. This commentary aims to explore the implications vW&R draw from their critique. In particular, it will raise objections to the moratorium argument and propose a presumptive case for commercializing AMAs.

The aim of this chapter is to shed new light on the question of what newly sighted subjects are capable of seeing on the basis of previous experience with mind- independent, external objects and their properties through touch alone. This question is also known as "Molyneux’s question." Much of the empirically driven debate surrounding this question has been centered on the nature of the representational content of the subjects' visual experiences. It has generally been assumed that the meaning of "seeing" deployed in these disputes is more or less clear and unproblematic, and therefore requires no analysis or clarification. In this chapter, we wish to challenge this assumption. We argue that getting clear on the meaning of "seeing" is the only feasible way to determine whether the empirical attempts to answer Molyneux’s question accurately capture what newly sighted subjects are in fact capable of seeing. Specifically, we show that the dominant interpretations of the empirical results from a recent study (Held et al., 2011) fail to take into account that seeing can be not only purely visual but also epistemic in that it requires background knowledge (such as what an object with a particular viewpoint-independent shape looks like from a particular perspective).