Search the archive.

AI & IP Policy: Regulating Disruptive Innovation. Intellectual property law has become the new battleground for ideas on how societies should deal with transformative change caused by disruptive tech.219 While trying to fill legislative gaps when laws cannot keep up with the pace of innovation, lawyers are becoming the policy makers. In this context, this section presents ten thoughts and policy suggestions on social, inclusive and innovation-friendly AI and data regulation.220 I start with a short legal policy vision of the direction I feel we should take. AI & IP for Dummies More or less protection for the owner of IP rights each has both advantages and disadvantages. More protection could stimulate costly and labor-intensive innova- tion, because that protection is offered as a reward. But the opposite, open access, also has social benefits. The degree of IP protection is therefore based on a consid- eration of pros and cons, and thus has a legal-political character. IP law policy aims to implement a regime that strikes a balance between un- derprotection and overprotection of IP rights—a regime that searches for an innova- tion optimum. That this is not an easy task as illustrated by IP history, which shows a pendulum swinging between underprotection and overprotection.221 Right now, we are in a stage of overprotection.222 In other words, intellectual property rights are getting too stretched. Besides that, IP law is poorly structured. This is not good because disruptive technologies such as AI, blockchain, and big data require a balanced, innovation-friendly regime. Solutions to resolve nega- tive effects that overprotection has are available. These include, on the one hand, technology-neutral open standards and exceptions or limitations such as the Ameri- can fair use principle, which allow more breathing room for both consumers and online platforms; and on the other hand, the introduction of an articulated public domain for AI Creations and Inventions (Res Publicae ex Machina). In general, we need more flexible IP laws and more open access. IP overprotection leads to market barriers for start-ups and SME’s and hinders international trade. When expressing preference for open access, AI & IP policy makers should ask themselves 2 questions: 1. To what extent can the assessment of the magnitude of the advantages and disadvantages of more open access versus more layers of IP rights be substantiated? This could, inter alia, be substantiated by sector specific empirical research. 2. And how and why do I weigh the pros and cons of this substantiation, and thus arrive at a preference for more open access? The goal should be at least a Pareto optimum/equilibrium, and preferably a Pa- reto improvement. B. 10 AI Related Policy Suggestions First of all, in order to have a sensible short-to-medium-term policy discussion about IP law in the context of emerging tech, it is important to demystify AI, resist anthropomorphisation and avoid speculation about the distant future. The state of the art is that we do not have Strong or General AI yet. What we have today is weak, pattern-based AI that is reaching task specific performance.223 It is a suite of tools that can be used for computer deductive reasoning and machine learning. Second, the unchartered terrain of IP and AI law makes it possible for legisla- tors to harmonize the acquis for AI internationally. For example, this can be done by introducing an official PD mark for AI Creations and Inventions (Res Publicae ex Machina). A harmonized, global acquis prevents forum shopping to countries such as the UK, Australia, and Japan and promotes legal certainty. It is vital that countries stop stretching IP rights at the expense of public domain and fundamental rights such as the right to information that are enshrined in the EU Charter and the USA Constitution. Third, AI related IP law policy should recognize the social value of disruptive technology and resist protecting settled market players who benefit from the status quo.224 IP law should not create barriers for new market entrants.

(Graham J. Reynolds University of British Columbia (UBC), Associate Professor, Faculty of Law, “The Precautionary Principle and its Application in the Intellectual Property Context: Towards a Public Domain Impact Assessment,” pg online @ https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2347922 //pa-ef)

While the factors noted above may already have been considered by bureaucrats and politicians in the legislative drafting process,58 several advantages flow from formalizing their consideration in a public manner through a PDIA. First, formalizing these factors and mandating their review through a PDIA would ensure that each factor is explicitly considered. Second, formalizing these factors and mandating their review through a PDIA would increase openness and transparency within the legislative process, ensuring that the factors are not simply analyzed by bureaucrats and politicians, but that they are seen, by the general public, to have been analyzed. Third, formalizing these factors and mandating their review through a PDIA would create a record of evidence that could be helpful for courts and others seeking to interpret the legislation. This record could also be relied on in future years in seeking to further reform, develop and shape intellectual property legislation and policy.

(Graham J. Reynolds University of British Columbia (UBC), Associate Professor, Faculty of Law, “The Precautionary Principle and its Application in the Intellectual Property Context: Towards a Public Domain Impact Assessment,” pg online @ https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2347922 //pa-ef)

Boyle defines the public domain as “material that is not covered by intellectual property rights.”9 For the purposes of this chapter, I will use Boyle’s definition as a starting point and expand upon it by suggesting that the public domain encompasses material that has never been covered by intellectual property rights; material formerly covered by intellectual property rights in which the grant of rights has expired (or has been declared invalid); uses of material that are deemed not to be covered by intellectual property rights through the application of defences/exceptions to intellectual property infringement (or that can be considered to be user’s rights10); and uses of material that are not covered by intellectual property rights by virtue of the application, by intellectual property owners, of flexible licences through which certain rights are disclaimed.11

(Graham J. Reynolds University of British Columbia (UBC), Associate Professor, Faculty of Law, “The Precautionary Principle and its Application in the Intellectual Property Context: Towards a Public Domain Impact Assessment,” pg online @ https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2347922 //pa-ef)

This is not to say that increasing intellectual property protection always (or ever) has a net benefit on the public domain. While increasing intellectual property protection may act as a further incentive for the creation of new expression, it also expands protection for existing works. Thus, although expanding intellectual property protection, for instance by increasing the term of copyright, may result in new works being created (making certain uses of those works, such as fair dealing, immediately available as part of the public domain), it also extends the period of time in which other uses of existing works will not be available as part of the public domain.

(Graham J. Reynolds University of British Columbia (UBC), Associate Professor, Faculty of Law, “The Precautionary Principle and its Application in the Intellectual Property Context: Towards a Public Domain Impact Assessment,” pg online @ https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2347922 //pa-ef)

A second option that could be pursued in the event that the PDIA process concludes that proposed legislative reforms might negatively impact the public domain beyond a certain threshold could be to require the legislative body to take steps to mitigate the potential impact of their proposals on the public domain. One step that could be taken is the creation of a body that has the authority to take certain steps to mitigate harm that might be done to the public domain should certain legislative reforms be enacted into law. This body could be given the authority to create or recommend the creation of additional defences to copyright infringement or to grant specific licences in order to partially (or completely) offset potential harm to the public domain that might result from the enactment of the proposed intellectual property reforms.60 Depending on how this mechanism is structured, democratic legitimacy concerns could arise under this approach, as well.

(Faculty of Law, University of Ottawa, “Evidence-Based Intellectual Property Policymaking: An Integrated Review of Methods and Conclusions,” pg online @ https://onlinelibrary.wiley.com/doi/full/10.1111/jwip.12069 //pa-ef)

Based on my analysis of the literature, and identification of its gaps and deficiencies, I identify a potential solution to the challenges of assessing the economic impacts of IP. I explore the usefulness of an innovation process framework as a practical means to structure the analysis of the socio-economic implications of IP policy decisions for specific sectors and across the economy. The framework explains how firms’ intellectual property management decisions at the microeconomic level are affected by IP policy decisions, and how firm level decisions relate to broader outcomes at the level of a particular industry or innovation system. Most importantly, this framework may help to shift the focus of policymaking from a firm-centric micro-economic perspective to a systemic macro-economic analysis. This is the crucial leap that few if any studies in the existing IP impact assessment literature have made. Such a framework could allow policy makers to investigate how the impacts of their decisions are likely to propagate through the economy, and who the winners and losers will be.

(Graham J. Reynolds University of British Columbia (UBC), Associate Professor, Faculty of Law, “The Precautionary Principle and its Application in the Intellectual Property Context: Towards a Public Domain Impact Assessment,” pg online @ https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2347922 //pa-ef)

Although recognizing that the risk of harm to the public domain that may flow from the expansion of intellectual property rights differs from the risk of harm to the environment that may flow from persons engaging in polluting or environmentally destructive activities, I suggest that it is appropriate to apply the precautionary principle in the intellectual property context. Both the environment and the public domain provide significant benefits to society. These benefits may be overlooked in favour of other benefits that might flow from development or enhanced intellectual property protection, respectively (such as economic benefits for certain industries or political benefits). As well, in the case of both the environment and the public domain, the impact of harm caused by polluting/environmentally destructive activities and intellectual property expansion, respectively, is both uncertain and difficult to establish.

(Graham J. Reynolds University of British Columbia (UBC), Associate Professor, Faculty of Law, “The Precautionary Principle and its Application in the Intellectual Property Context: Towards a Public Domain Impact Assessment,” pg online @ https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2347922 //pa-ef)

D. Conclusion

(Graham J. Reynolds University of British Columbia (UBC), Associate Professor, Faculty of Law, “The Precautionary Principle and its Application in the Intellectual Property Context: Towards a Public Domain Impact Assessment,” pg online @ https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2347922 //pa-ef)

Fourth, and the subject of this chapter, another situation in which the precautionary principle could be usefully applied is at the point at which proposals for intellectual property reform are introduced. The application of the precautionary principle in the context of intellectual property reform could require legislative bodies, when evaluating proposed amendments to their intellectual property legislation or new intellectual property legislation, to explicitly consider the impact of any such proposals on the public domain. In the following section, I will consider one possible instantiation of the precautionary principle in the context of intellectual property reform, namely in the form of a PDIA. 3) Applying the Precautionary Principle in the Context of Intellectual Property Reform: Towards a Public Domain Impact Assessment In considering how the precautionary principle might be applied in the context of intellectual property reform, it is informative to look to existing works that have advocated for an approach to intellectual property reform that could be characterized as “precautionary.” One commentator, Thomas F Cotter, has explicitly suggested the application of the precautionary principle in the context of copyright, stating that “policymakers would be wise to incorporate something analogous to the Precautionary Principle, in order to minimize the risk that aggressive copyright laws will decimate the cultural environment.”46 Cotter proposes several ways in which the application of the precautionary principle could impact intellectual property reform. First, he suggests that the principle could “shift the burden of justifying a proposed, but potentially harmful, rule to the affected industry.”47 Cotter states that the burden of proof could vary depending on factors such as the “magnitude of the potential harm, its irreversibility, and the current state of scientific understanding of its probability.”48 Second, Cotter suggests that “advocates of further copyright expansion” could be required to demonstrate, through “some meaningful degree of proof that further expansions are necessary to maintain incentives, and are likely to do no harm to the other relevant goals of copyright.”49 Modifying the intellectual property reform process in such a manner as to require evidence demonstrating the probability and seriousness of harm that may result to the public domain from the adoption of proposed reforms, and demonstrating the consistency of proposed reforms with the goals of intellectual property laws, could be of significant benefit to the public domain. It is unclear, however, how the burden shift suggested by Cotter might function should multiple industries propose the same reform or should reform be proposed by parties other than industry. As well, questions could be raised as to whether it is in the public interest to permit the industry proposing legislative reform to act as the party providing evidence justifying this same reform. Cotter, acknowledging the potential for abuse inherent in such an approach, notes that “affirmative findings [could be required] from the Copyright Office or from Congress.”50 Building on Cotter’s approach, I will conclude this chapter by proposing another possible instantiation of the precautionary principle in the context of intellectual property reform. Specifically, I will propose the creation of a PDIA. I envision the PDIA as a process through which proposals for intellectual property reform, prior to their enactment, are evaluated by an independent review panel in order to determine their potential impact on the public domain. Given the space constraints of this edited text, I will not describe the PDIA in exhaustive detail. Instead, I will introduce the framework of the PDIA, leaving the specific details to be expanded upon in a future work. My proposed PDIA is modeled on the Canadian Environmental Assessment Act (CEAA).51 The CEAA can be seen as an instantiation of the precautionary principle in the context of Canadian environmental law. As in the case of intellectual property, in which private rights are balanced with the public interest, the CEAA attempts to achieve a balance between economic development and environmental concerns. I envision the PDIA process as an open, public process, conducted by an independent review panel. Individuals, groups, and industry could have the opportunity to submit documents and provide oral testimony to the review panel with respect to the impact of certain legislative proposals on the public domain. Documents and testimony under consideration by the panel could then be published online, giving the public the opportunity to examine and comment on the submissions, the evidence contained in the submissions, and the methodology employed by parties that have submitted evidence.52 Structuring the process in such a manner, as opposed to relying on the party desiring a specific reform to provide evidence justifying that reform, reduces the risk that the party desiring the specific legislative reform might, as Cotter states, make “rosy predictions” or engage in “sleight-of-hand.”53 As is the case in the procedure set out under the CEAA, a number of mandatory factors could be considered under the PDIA process, including:54 - The projected impact of the proposed legislative reform on the public domain, including the cumulative effects on the public domain that are likely to result from the proposed reform in combination with existing legislation or other reforms that have been or will be undertaken. - The significance of this impact. - Comments from the public received during the course of the PDIA process. - Measures that might mitigate any adverse impacts of the proposed legislative reform on the public domain.5 Evidence demonstrating that the proposed reform will (or will likely) achieve the desired result.56 - The impact that the proposed reform might have on groups.57 While the factors noted above may already have been considered by bureaucrats and politicians in the legislative drafting process,58 several advantages flow from formalizing their consideration in a public manner through a PDIA. First, formalizing these factors and mandating their review through a PDIA would ensure that each factor is explicitly considered. Second, formalizing these factors and mandating their review through a PDIA would increase openness and transparency within the legislative process, ensuring that the factors are not simply analyzed by bureaucrats and politicians, but that they are seen, by the general public, to have been analyzed. Third, formalizing these factors and mandating their review through a PDIA would create a record of evidence that could be helpful for courts and others seeking to interpret the legislation. This record could also be relied on in future years in seeking to further reform, develop and shape intellectual property legislation and policy. The consequences of determining, through the PDIA process, that proposed legislative reforms might negatively impact the public domain could vary. One possible approach could be for legislative bodies to bind themselves to the determination arrived at through the PDIA process.59 This approach would represent the strongest instantiation of the precautionary principle of the options outlined in this chapter. Under this approach, if the PDIA process concludes that proposed intellectual property reforms might negatively impact the public domain beyond a certain threshold (for instance, might have a “severe impact” on the public domain), the legislative body would not be permitted to pass the legislation as proposed.

The notions of competency and capability have long been used as a means of excluding disabled people from a range of different settings including education, work and leisure activities, as well as denying them access to opportunities such as relationships, parenting and friendships (Goodley Citation2014, 156). The response of much of postconventional disability studies has been to analyse the cultural origins of this exclusion, but whilst this genealogy may explain why this has happened and explore the ideological origins, it fails to unpack carefully why they are wrong. Goodley acknowledges that there is a need for a normative stance but argues that calls for normative judgments run the risk of ‘leaving dis/ability only in the realm of normative [sic] … [and] ignoring the productive potential of dis/ability to crip a whole host of norms that are, in actuality, limiting and stifling’ (Citation2014, 158). These, he suggests, include characteristics such as rationality, capability, responsibility and competency; which he identifies as ableist ideals, part of the modernist humanistic subject (Citation2014, 156). This raises a number of questions such as whether rationality and responsibility are normatively on a par with irrationality and irresponsibility. Surely not. Homophobia, for example, is an irrational and morally harmful mindset. It is based on irrational fears about contagion, divine retribution and so forth, and results in moral harms to sexual minorities such as discrimination, hate speech and hate crime.

How about (ir)responsibility? Imagine Bill is by himself taking care of his infant child at his family’s remote country house and all of a sudden feels an irresistible urge to drink three bottles of wine. Getting intoxicated is, perhaps, a perfectly legitimate interest as such, but in those circumstances it would compromise Bill’s capability to properly supervise his child and thus put the child at risk and in danger. In other words, rationality and responsibility in general are good, irrationality and irresponsibility are bad.

To understand more fully how activists might productively engage with and respond to failure, I turn to the role of failure in utopian political thought and action. Utopia has traditionally been conceived as a literary and philosophical tradition dedicated to developing a blueprint of a perfect society. In the last century, critical theorists, from Walter Benjamin to Theodor Adorno, and poststructuralist thinkers such as Michel Foucault have helped rethink utopianism as part of an effort to problematize Enlightenment faith in progress (van der Walt 2014). Ernst Bloch in particular has become known for his broader understanding of utopia which includes a wide array of “dreams of a better life”, expressed in anything from architecture to music. Utopias provide less of a universal blueprint than an “expression of hope” that acts as a “directing act of a cognitive kind” (Bloch 1995, 1, 12). His heirs have articulated two key elements to this conception of utopia that are crucial for thinking about the failure of the ADA and transformative law: its ambiguity and provisionality.

Utopias are ambiguous because they constitute partial, limited attempts to imagine a better world which are formulated using the language and frameworks of the present. Utopias simultaneously present a moment of radical disjuncture to think and act freely using our imagination, but also “force us to face the fact that we do not live there; we live here, and we cannot but use the language of the here and now in all our imaginings” (Winter 2006, 3). The question then is how to respond to this ambiguity: do we presuppose that we have the means to identify our vision of the world as the only correct one and to resolve all problems that come with implementing this vision in its “true” form, or do we affirm the limitations of our ability to imagine the not-yet.

Moshe and Linda Steele,” feminists@law, Vol 12, No 1 (2023), dml)

In closing, I turn to the remedial pathways contemplated by the authors. Both books are written from a place of hope – the hope that the carceral logics shoring up capitalist, colonialist and ableist power may be radically subverted, not just superficially reformed or diverted. Yet rather than giving readers (including lawyers) specific instructions, the authors offer general observations aimed at generating new ways of thinking, working and being together.

One of the most famous centers was founded in Berkeley by students under the leadership of Ed Roberts (Levy, 1988). Roberts had successfully litigated his way into further education at the University of California, Berkeley, where he was influenced by the counterculture of the 1960s. The aim of the center was to serve the local community of disabled people, but it also played an influential role at the beginning of the disability rights movement, including in funneling federal and local funds towards supporting disabled people and activism and in forming a normative vision of a better society (Charlton, 2000, p. 132). The center offered one, important, launchpad for the leadership of the disability rights movement to launch protests, inform wider society about the efforts of the movement and changes to legal provisions for disabled people, and coordinate with other groups to develop and lobby for new legislative initiatives. The independent living centers are thus practical criptopias that used the material and ideational structures of their time (such as federal and local funding and consumer ideology) to produce concrete evidence that—with suitable community support for people with physical impairments to overcome social and physical barriers, such as inaccessible housing—a fully integrated society is possible and plausible.

Secondly, the movement pursued a grounded utopia by lobbying for the passing of legislation to enforce equal status as citizens and contributors to the economy. The focus on law may seem antithetical to the utopian project because laws serve to regulate conflict and disagreements which arise in imperfect societies, for instance about the distribution of limited resources. For this reason, classic writers such as More highlighted the absence of a comprehensive body of laws as a notable feature of their utopian society (Herman, 2016; More, 1685, p. 148). I introduced above how utopianism has since broadened beyond the articulation of universal blueprints of the perfect society and this move necessitates a greater concern with how to transform an unjust legal system. A broader conception of utopia also reveals the social dreaming that motivates legislation, most obviously in the articulation of, and struggle for the protection of, human and civil rights (cf. Moyn, 2012, p. 1).

Scholars working in this intellectual tradition typically insinuate how things, as they currently stand, are wrong whilst providing very little practical ethical guidance as to how things ought to be (Sayer Citation2011, 44 and 229; Vehmas and Watson Citation2014). Whilst postconventional disability studies clearly is an ethical and political project, exemplified in many recent works (for example, Goodley, Lawthom, and Runswick-Cole Citation2014; Shildrick Citation2012), these accounts lack an engagement with the normative issues of how things ought to be. Current postconventional normative analysis does not properly articulate the moral and political wrongs disabled people face, and neither does it give us practical guidelines regarding ways to right those wrongs (cf. Lundblad Citation2011, 146).

Goodley and Runswick-Cole (Citation2015), for example, have recently produced an analysis of Big Society ideas in the United Kingdom and how the present neo-liberal agenda has damaged disability services and disabled people’s social status in general in the United Kingdom. They conclude that we should question ableist norms guiding social arrangements, and that we should ‘continue to question, destabilise assumptions that marginalise and exclude bodies and minds that are judged to fail to meet the expectations of ableist normativity’ (Goodley and Runswick-Cole Citation2015, 10). Using the schema developed by Fraser (Citation2013), we would argue that there are two clear claims in the arguments contained in this and other works in postconventional disability studies (see, for example, McRuer Citation2013; Wolbring Citation2008). Firstly, disabled people emerge via regulatory systems and forms of normality, and it is only in these highly ableist regulatory schemas that the contemporary disabled subject is constructed. Disabled people become constituted through the power/discourse formations where all subjectivities are always already culturally constructed (Fraser Citation2013). There is in essence no doer behind the deed.

My discussion of these three styles of ‘doing’ normativity has pointed at three problems disability scholars run into in their normative analyses. Let me briefly list them in turn. The first problem resides in the ‘top-down’ style of ethical analysis found in the materialist and critical realist approaches. By ‘applying’ moral principles to arbiter social practices, such analyses threaten to render moral theory static and irresponsive to the complexity of real-life moral tensions and dilemmas, and also leave little room for forms of normativity internal to practice. (Vehmas and Watson’s suggestion to engage with normativity through the analytical philosophical method seems to imply a similar ‘top-down’ style of analysis; hence my reservations about their proposal.) The second problem resides in the separation of ‘fact’ and ‘values’. I traced this problem explicitly in the critical realist approach, but it is in fact not considered carefully in any of the approaches. Here, the problem is that scholars are insufficiently aware that normativity is always already a part of the empirical reality we observe—a point that, if we want to ‘do’ normativity right, ought to be explicated and explored. The third problem has to do with the sort of ethical directives these approaches produce. I noted of the cultural approach that its suspicion of normativity does not allow for constructive normative conclusions, meaning that it has trouble providing meaningful normative guidance. By contrast, the opposite seems true for the materialist and critical realist approaches; their rigid use of moral principles makes it unclear to what extent these approaches can generate normative insights that are sufficiently complex, nuanced, and in tune with practice—and thus, provide insights that can actually help improve the lives of disabled people. In effect, disability scholars are torn between giving too much or too little in the way of normative directives.

My overarching concern in signaling these problems is that the conventional styles of ‘doing’ normativity in disability studies might fall short of what I believe should be their aspiration: that is, to generate normative insights that speak to the nuances and complexities of the lives of disabled people—and hence, that might help to improve these lives in practice. On an analytical level, these problems spring forth from the interaction between ethical ‘theory’ and empirical ‘data’ and the sort of conclusions this interaction can generate. This issue is also at the root of my hesitation regarding the method proposed by Vehmas and Watson (Citation2016), which in my view does not say enough about how empirical knowledge ought to inform moral reasoning.

The sheer volume of data that the military collects is overwhelming and requires the help of computer analysis. Running an algorithm is a much more effective way to make sense of all the raw data. It reduces big data into a digestible amount of information that a human can understand. This is how military leadership can make actionable decisions based on facts. The military can make tremendous improvements with their strategic decision-making, including quick and exact scenario analyses. During urgent circumstances, AI can neutralize biases. Artificial intelligence is a prime tool for training, lending itself to opportunities to give soldiers and personnel individualized instruction. Computer models and technology are reliable. Simulations and exercises of the future are becoming more and more real, helping the military better prepare for times of warfare. AI can bridge gaps in military defenses, making departments less fragmented. Data processing runs more efficiently from various sources. Logistics is more proactive because it can make administrative tasks easier. Thanks to autonomous logistics, troops are in less danger, and support systems are more enhanced.

The battlefield is often a location that is extremely hazardous, and AI may mitigate some threats that the military faces. A prominent example of how AI is helping to keep soldiers safer during times of battle is that soldiers can use artificial intelligence to identify dangers faster, and AI can provide soldiers guidance on how they can handle potential dangers. Today’s ground vehicles and army aircraft have intelligent sensors. There is less need for humans, alleviating risk to human life.

Rescuing wounded soldiers during active battlefield situations is one of the most dangerous positions and the cause of many military deaths. According to researchers, around 86% of battlefield deaths occur during the first thirty minutes post-injury. This is the reason researchers across the world are working toward developing artificial intelligence (AI) technology that can bolster battlefield healthcare.

Artificial intelligence (AI) can reduce serious injury and death on the field of battle. Most of the development is occurring in the application of robotic systems and unmanned ground vehicles (UGVs) to implement and provide aid in on-site surgery; look for and perform casualty extraction; and conduct intelligence, surveillance, and reconnaissance (ISR) activities.

One of the best benefits of artificial intelligence in the military is in an area that military commanders might feel hesitant to let AI contribute. That is helping with strategic decision making. AI’s algorithms are able to collect and process data from numerous different sources to aid in decision making, especially in high-stress situations. In many circumstances AI systems can quickly and efficiently analyze a situation and make the best decision in a critical situation. It is also able to neutralize prejudices that may come with human input, with the caveat that AI may not yet have a fully developed understanding of human ethical concerns and there is a danger of AI learning from the biases that may exist in materials in its database. However, decision making under pressure is a critical part of being a service member, and AI and humans can work together to make this process easier. The combination of humans’ ethical understanding and AI’s quick analytical abilities can speed up the decision making process. Generative AI can contribute to the decision-making process in military settings. Rapidly sorting through large amounts of data, generative models can show connections, patterns, and potential implications that humans alone would take a longer time to find. This information can be presented to human decision makers not only as reports but also in a conversational format, facilitating human-AI collaboration. AI can also create simulations to test out possible scenarios, allowing for more informed decision making. Upon receiving this information from AI, humans must fill in the gaps, using their understanding of ethical principles, national security interests, and situational nuances to create optimal outcomes.

The United States government is leading global efforts to build strong norms that will promote the responsible military use of artificial intelligence and autonomous systems. Last week, the State Department announced that 47 states have now endorsed the "Political Declaration on Responsible Military Use of Artificial Intelligence and Autonomy" that the government first launched at The Hauge on Feb. 16.

AI refers to the ability of machines to perform tasks that would otherwise require human intelligence, such as recognizing patterns, learning from experience, drawing conclusions, making predictions or generating recommendations.