Free beta access. 50 free AI tokens each week. Optional top-ups available. See what is included
← Back to archive
2022 Michigan Affirmative

Aff - AI Ethics - Starter - Michigan 7 2022

Aff - AI Ethics - Starter - Michigan 7 2022.docx

Download Word file
Indexed text preview
Aff – AI Ethics Aff – UM22 Starter Packet
Starter Notes
Brief Explanation
The militaries of the world are incorporating Artificial Intelligence into their weapons, planning, operations, logistics – pretty much everywhere. Some countries are going all-in – developing Lethal Autonomous Weapons (LAWs - killer robots) and nano-drone swarms, known charmingly as Genocide Swarms. Others are more cautious and are limiting their AI use to wargame simulations, or targeting mods. Perhaps the biggest issue in controversy is the level of Autonomy. Fully autonomous weapons exclude humans from the kill chain – search, target acquisition, engagement, execution. Low Autonomy weapons might be an AI adversary in a simulated attack against Russian forces in Ukraine. Very few states have (allegedly) developed or deployed fully autonomous weapons, but as the technology improves exponentially, that is a possible future.
This affirmative plan calls for human control – a “human in the loop” who has final say over AI functions. It has the US engage NATO in establishing ethical principles, known as RAI (responsible AI use), for all NATO operations and weapons. There are three reasons for this:
Cohesion. If everyone in NATO is on the same page and coordinated on technology and tactics, that is called Interoperability. The command, political, and public will to get coordinated is called Cohesion. Interoperability and Cohesion are the foundation of effective coalition military forces and operations. AI threatens to disrupt that cohesion and interoperability. If the US has Fully Autonomous weapons, and Italy does not, then those weapon systems will have trouble interoperating (not sure if that is a word). If the US is using mostly automated systems, and France is sending mostly human troops, then France might ask why they should risk French lives when the US is not. If the US spends A Lot on Fully Autonomous weapons, but Lithuania doesn’t have that technology, then the US might feel that other nations are not sharing the burden. If the British public opposes LAWs, and they are in a conflict allied with the US, who is using LAWs, then the British public might demand that their government stop supporting the operation. Or the US might perceive that the UK would stop supporting it. Dialogue to cooperate to establish norms for AI throughout the alliance would improve cohesion and interoperability, which makes for a more effective and credible NATO, which stops a nuclear war.
Crisis Instability. Many of the new AI weapons are “brittle” – they tend to react poorly when the situation goes outside of their programmed parameters. Like if the Ukrainians put reflective tape on stop signs to disorient the Russian drones piloted by AI. For example, without a human in the loop, AI accelerates battle decisions to machine speed. If something goes wrong, it goes Really Wrong Really Fast, without anyone able to check the escalation to nuclear war. Back in the 80’s, Russian Early Warning Radars detected what it thought to be a US nuclear attack and signaled that to Soviet Rocket Command. But a Lieutenant Colonel – Petrov – trusted his gut instinct and said “Maybe is birds, comrade.” Crisis averted. AI would have sent the nuclear weapons in a split second. There are many crises in the future for Europe, and miscalculation or instability in them could escalate a conflict accidentally or mistakenly. Human control would provide a stop in the escalation chain to prevent war.
Human Dignity. Autonomous AI weapons would make decisions about life and death by algorithm. The soldiers who are killed would just be part of a decision tree – depersonalized and objectified. Civilian casualties would become data points. This is unethical because it does not respect human dignity. Now, to be clear – all war is undignified. A human soldier shooting you leaves you just as dead as an autonomous drone. However, that soldier shares your humanity. They feel compassion, and hope, and empathy, and regret. They know what it means to take a life because they value their own. AI has none of that – it cannot “know” what it means to kill or die, because it cannot “feel” life. For better or worse, if war involves lethal decisions, the choice to kill another person should be made by a person. International and Human Rights laws are founded on the principle of Human Dignity. Our actions, arguments, motives and intentions must respect the Dignity of all persons. Death by algorithm does not respect the dignity of individual people, because AI cannot “know” what it means to be alive. (This is where the Cybernetics K comes in…) This is a deontological argument, not a consequential or utilitarian one. It says that before you decide if plan has good or bad outcomes, you first have to determine whether it is Right or Wrong. If AI weapons are unethical, you never even evaluate if they have good consequences – that is the Ethics First framework argument.
Some Comments
The 1AC is WAY too long to read all three advantages – it is impossible to run them all. DON’T destroy the quality of the evidence by trying to highlight it down to read all three. Instead, choose two, or maybe even one, advantage to read, and cut out cards before highlighting into incoherence. For instance, if you want to run Human Dignity, that works pretty well all by itself.
I know that there are three inherency cards and like 8 extension or AT blocks on inherency, and that is too much. In fairness, this case is probably not inherent, so “Be Prepared!”
A lot of the literature and international debate is about LAWs. It’s more exciting to talk about Killer Robots. However, that is only a small fraction of the military uses for AI in NATO, so don’t get hung up on that. On the other hand, “human control” means that weapons are no longer “autonomous” – at least not fully autonomous. So the plan does de facto ban LAWs.
1AC
1AC - Inherency
Contention One – Inherency – The Pentagon does not require Human Control for artificial intelligence in weapons.
Allen, 2022 - Director, AI Governance Project, Strategic Technologies Program at CSIS [Gregory C. June 6 “DOD Is Updating Its Decade-Old Autonomous Weapons Policy, but Confusion Remains Widespread” https://www.csis.org/analysis/dod-updating-its-decade-old-autonomous-weapons-policy-confusion-remains-widespread Acc 6/6/22 TA]
In November 2012, the Department of Defense (DOD) released its policy on autonomy in weapons systems: DOD Directive 3000.09 (DODD 3000.09). Despite being nearly 10 years old, the policy remains frequently misunderstood, including by leaders in the U.S. military. For example, in February 2021, Colonel Marc E. Pelini, who at the time was the division chief for capabilities and requirements within the DOD’s Joint Counter-Unmanned Aircraft Systems Office, said, “Right now we don't have the authority to have a human out of the loop. Based on the existing Department of Defense policy, you have to have a human within the decision cycle at some point to authorize the engagement." He is simply wrong. No such requirement appears in DODD 3000.09, nor any other DOD policy. Misconceptions about DODD 3000.09 appear to extend even to high-ranking flag officers. In April 2021, General Mike Murray, the then-four-star commander of Army Futures Command, said, “Where I draw the line—and this is, I think well within our current policies—[is], if you’re talking about a lethal effect against another human, you have to have a human in that decision-making process.” Breaking Defense, a news outlet that reported on Murray’s remarks at the time, stated that the requirement to have a human in the decisionmaking process is “official Defense Department policy.” It is not. DODD 3000.09 does not ban autonomous weapons or establish a requirement than U.S. weapons have a “human in the loop.” In fact, that latter phrase never appears in DOD policy. Instead, DODD 3000.09 formally defines what an autonomous weapon system is and requires any DOD organization proposing to develop one to either go through an incredibly rigorous senior review process or meet a qualifying exemption. Regarding the latter, cyber weapons systems, for example, are exempted. The DOD recently announced that it is planning to update DODD 3000.09 this year. Michael Horowitz, director of the DOD’s Emerging Capabilities Policy Office and the Pentagon official with responsibility for shepherding the policy, praised DODD 3000.09 in a recent interview, stating that “the fundamental approach in the directive remains sound, that the directive laid out a very responsible approach to the incorporation of autonomy and weapons systems.” While not making any firm predictions, Horowitz suggested that major revisions to DODD 3000.09 were unlikely.
Our military does not prioritize ethics in the deployment of AI.
Lawrence and Cordey, 2020 – researchers for The Cyber Project at the Belfer Center for Science and International Affairs [Christie and Sean, August, The Cyber Project Paper “The Case for Increased Transatlantic Cooperation on Artificial Intelligence Edited by Lauren Zabierek and Julia Voo https://www.belfercenter.org/sites/default/files/2020-08/TransatlanticAI.pdf Acc. 4/21/22 TA]
In recent months, Head of Global Governance, Regulation, Innovation and the Digital Economy at the Centre for European Policy Studies Andrea Renda pointed out that interests and political support for greater transatlantic coordination and collaboration on AI seems to be increasing. This is notably demonstrated by the visit in early 2020 of high-level officials on both sides of the Atlantic. The first, as already mentions in the Political collaboration section, was the visit in January from former Director of the JAIC Lt. Gen. Jack Shanahan and then CTO Nand Mulchandani to Brussels. During their visit to the NATO headquarters and with EU leaders, Lt. Gen. Shanahan underlined the need for greater collaboration for AI, particularly within NATO, in order to counter revisionist states such as China and Russia that promote digital authoritarianism. As Acting Director of the JAIC Nand Mulchandani mentioned, Lt. Gen. Shanahan also underscored the need for the US and Europe to rally around common values and not allow technicalities around ethics to block collaboration. 279 The second high level visit was by a delegation from the European Parliament’s Civil Liberties Committee, which visited Washington D.C and Boston in February 2020. The delegation held meetings with US Congressional members and representatives across the US Administration, including the DOS, DOJ, DHS, FTA, FBI, and PCLOB. The Members of the European Parliament (MEPs) also held meetings with representatives from industry, universities, think-tank, start-ups, and NGOs. The discussions centered on a variety of AI topics, ranging from potential future US federal legislation on personal data protection, the use of artificial intelligence in law enforcement, and the visa waiver program.280
NATO states have differing positions on responsible use for AI – the potential exists for agreement, but states are implementing their own policies.
Stanley-Lockman, 2021 - Center for Security and Emerging Technology [Zoe CSET Issue Brief August “Responsible and Ethical Military AI Allies and Allied Perspectives” https://cset.georgetown.edu/publication/responsible-and-ethical-military-ai/ Acc. 4/21/22 TA]
With more of a focus on norms and state responsibility, a broader definition of responsibility beyond the DOD “responsible AI” principle can also introduce new convergences for U.S. RAI implementation. The DOD RAI Strategy and Implementation Pathway recently tasked by Deputy Secretary Kathleen Hicks can incorporate these views and help allies refine their approaches to AI ethics in order to enable greater cooperation and allied AI adoption. This is because many allies see firm approaches to managing ethical risk as a prerequisite policy question before investing in AI-enabled capability development—including defensive systems and countermeasures. Overall, international engagement is mutually beneficial to responsible AI endeavors. The United States should look at how other countries are implementing their approaches, just as the United States can exert influence and maintain its leadership role in responsible and ethical AI for defense by helping its allies and partners form their own views in alignment with one another. Conclusion No single actor has a monopoly on the answers to implementing responsible AI in any high-risk area, let alone defense. Cooperation is therefore important to collectively navigate the difficulties of responsible governance of emerging technologies. For DOD, the focus has been predominantly on the transposition of safe and ethical AI principles into action. Rather than adopting principles for defense, some allies are moving straight into implementation. Thus far, this is borne out in tabletop exercises, outreach, ministerial committees, ethical reviews, education and certifications, exercises and trials, and defense programs of record. It is too early to judge these fledgling efforts, but tracking their evolution may prove useful to broader AI ethics implementation, be it for other defense ministries or even civilian actors. While jumping straight to implementation can mean a more pragmatic focus on tools, tracking how different AI stakeholders use those tools may be more difficult. In this way, principles can be seen as a helpful organizing force, as is the case for DOD. This said, the scale of the U.S. military bureaucracy and national security innovation base may require higher visibility relative to allied counterparts. Still, another key difference is precisely this visibility. The analysis here is based on information in the public domain, which may also partially explain its transatlantic tilt. The U.S. approach to responsible and ethical AI for defense also differs from other countries in that the consultation and process that led to its principles is far more transparent than is true for most allies. A possible Catch-22 could be at play here, with allies reticent to publicize approaches to such controversial issues, despite the fact that offering such inroads can build trust and confidence that governments are handling these high-stakes questions responsibly.
1AC - Plan
Plan: The United States federal government should substantially increase its security cooperation with the North Atlantic Treaty Organization on ethical principles to ensure human control in artificial intelligence systems.
1AC - Advantage – Cohesion
Advantage – NATO Cohesion
NATO does not have a coordinated position on the ethics of AI – this undermines military and political cohesion, which undermines NATO militarily.
Stanley-Lockman, 2021 - Center for Security and Emerging Technology [Zoe CSET Issue Brief August “Responsible and Ethical Military AI Allies and Allied Perspectives” https://cset.georgetown.edu/publication/responsible-and-ethical-military-ai/ Acc. 4/21/22 TA]
At the same time, significant differences in ethical approaches to AI in defense could imperil political cohesion and undermine coalition success. Politically, alignment on ethics is important because shared values are at the foundation of U.S. alliances.3 This also trickles down to the operational level, where differing views on ethics could mean that allies field their systems with different legal authorizations and rules of engagement.4 If coalition partners deem each others’ capabilities to be based on different legal, ethical, and doctrinal assumptions, then forces may not be able to communicate and operate together.5 Further, if different ethical bases for capability development mean that some countries have higher thresholds for what they develop and contribute to coalition operations, then others may perceive them as not equally sharing risks to life.6 As such, political cohesion and policy considerations about ethics could directly influence operational effectiveness. In other words, failure to align allied perspectives on AI ethics in defense will inevitably undermine the ability of allied forces to understand each other and work together.7
Diverse AI standards undermine military interoperability because it prevents different nations militaries from cooperating. NATO dialogue is key to coordinate ethical standards for military AI. Meaningful human control which respects human dignity is essential to political cohesion.
van der Merwe, 2021 - Fellow Center for European Policy Analysis Defense Tech Initiative [Joanna, Feb 17, “NATO Leadership on Ethical AI is Key to Future Interoperability” https://cepa.org/nato-leadership-on-ethical-ai-is-key-to-future-interoperability/ Acc 4/16/22 TA]
In October 2020, Deputy Secretary General of NATO Mircea Geoană highlighted the benefits of establishing a “transatlantic community cooperating on Artificial Intelligence (AI).” The Deputy Head of NATO’s Innovation Unit followed with a commitment to its responsible use. The US Department of Defense (DoD) adopted Ethical Principles for AI in 2020 and has committed to bringing together NATO member and partners to operationalize these principles. Despite these statements and developments, more work is required to tackle the very real challenge that ethical AI will pose to future interoperability within NATO. Without a NATO-led initiative focused on aligning these ethical principles across the Alliance, the interoperability risk of nations fielding AI-based systems that hinder joint operations is high. As the foremost security framework for Europe and North America, as well as the leading defense alliance for promoting and protecting democratic values, NATO is able to facilitate alignment on this issue. As part of a broader strategy on emerging and disruptive technologies, NATO must prioritize ethical AI if it wishes to promote the shared values upon which it was founded, play a key role in facilitating innovation across the Atlantic, and ultimately retain the ability of its members to undertake joint operations. Establishing NATO ethical AI principles is the first step toward both technical and political alignment, in turn enhancing and fostering interoperability, which is the foundation for NATO to respond to emerging threats as an Alliance, in a flexible and timely manner. A key challenge for NATO is raising awareness that the answers to ethical questions can no longer be left to later stages of the development and procurement cycle. Decisions made at the political and legal level will have a significant impact on the engineering practices used to develop AI, as well as the technical characteristics of the AI-based systems. The answers to questions such as respecting human dignity, human control, and accountability will be the foundation upon which many technical elements are programed. Systems developers need to make a number of calls throughout the development cycle informed by the answers to key questions, including: how to label data what data to use, and what is an acceptable outcome? These answers will also impact how AI systems are evaluated and ultimately deployed. If individual nations or groups are left to develop their own ethical principles without wider alignment to NATO, the result will be a number of AI-based systems with varying technical specifications based on the legal and policy decisions made by individual governments when answering the key questions. As has been demonstrated in areas such as facial recognition and policing algorithms, the assumptions made by those developing the tools and answering the key questions have a significant impact on the real-world functioning of the tool and societal acceptance of its ethics. The risk of tools failing to gain acceptance depends on the legal and ethical decisions made by governments. For the military, this may mean one state using an AI-based system that is seen as unacceptable by another, and in a joint oper

Preview shortened. Download the file for the full document.