29 November 2024
IP24100 | Can Moratoria Break the Current Arms Control and Disarmament Deadlock?
SYNOPSIS
At a time when the prospects of concluding new arms control and disarmament agreements are not good, it is important to consider other tools that could help. One such tool is a moratorium — a state’s voluntary suspension of specific military activities. While there were some limits to the effectiveness of moratoria in the past, they may still be useful in the current context, where there is a lack of agreement on topics such as the use of AI in the military domain.
COMMENTARY
On 16 November 2024, presidents Joe Biden and Xi Jinping held a bilateral meeting in Lima. The two leaders agreed on the need for prudence and a sense of responsibility in developing AI technology in the military field. They also affirmed the need to maintain human control over the decision to use nuclear weapons. Although made in the context of a bilateral meeting, such an affirmation is a form of moratorium. Many want to see all nuclear weapons possessors make such a declaration.
On 21 November, Russia launched an intermediate-range ballistic missile on the Dnipro region in Ukraine. One reason for the strike was to respond to Ukraine’s long-range missile strikes on Russia, only days after their use had been authorised by the United Kingdom and United States. Another reason was Washington’s withdrawal from the Intermediate-Range Nuclear Forces (INF) Treaty and its subsequent refusal to agree to Moscow’s proposal to declare a moratorium on the development of missiles previously banned under the treaty. Had such a moratorium been in place, the Dnipro ballistic missile strike may not have happened. Russia has also recently stated that it would revoke this moratorium if the United States went ahead with plans to deploy short- and medium-range missiles to Europe and Asia.
What Is a Moratorium and Why Are Moratoria Used?
A moratorium, in the context of arms control and disarmament, is a voluntary suspension by a state of specific military activities, such as the testing, production, or deployment of weapons systems. Moratoria are politically binding on the countries that issue them. They are not legally binding, but most states, once politically bound, consider moratoria to have the same effect as a legal obligation.
States can use moratoria proactively to signal that they have decided to stop doing a certain thing and put pressure on others to make a similar commitment, which could then be turned into a legally binding agreement. A long-standing example in the space domain is the Russian initiative on “No First Placement of Weapons in Outer Space”. A more recent example is the US moratorium on direct-ascent anti-satellite testing (DA-ASAT).
A moratorium can also be an important political signal to create the environment for negotiating new legally binding instruments. In such cases, it demonstrates a state’s willingness to enter negotiations in good faith. An example is the moratorium announced by the “Permanent Five” UN Security Council members (P5) on nuclear testing, which preceded negotiations for the Comprehensive Nuclear Test Ban Treaty (CTBT).
Even when starting negotiations is unlikely, moratoria can still be useful as confidence-building measures. For example, France, Russia, the United Kingdom, and the United States have moratoria on the production of fissile material for use in nuclear weapons. However, their commitments have failed to persuade China and Pakistan, which do not have moratoria, to enter negotiations for a Fissile Material Cut-off Treaty (FMCT). India, which is in favour of an FMCT, does not see the value of a moratorium in persuading these two to come to the table.
China has adopted a similar approach with its policy of “No First Use of Nuclear Weapons”. Its undertaking that it will not be the first to use nuclear weapons is a form of moratorium and is intended as a first step towards negotiating its “Draft Treaty on No-First Use of Nuclear Weapons”.
Moratoria have also been suggested as a means of retaining some elements of a legally binding instrument that has fallen away. After Washington withdrew from the INF Treaty, citing Russian violations, Moscow proposed a moratorium on the development of missiles previously banned by the treaty.
Are Moratoria Effective?
Moratoria are voluntary — they can be revoked at any time. For some, this reality means moratoria have little value. On the contrary, the longer a moratorium is in place, the longer it has been adhered to, meaning it does have value. A good example here is the moratorium on nuclear testing.
However, that leads to the next criticism of moratoria, namely, that they are not verifiable and therefore there is no guarantee that countries which have issued them are abiding by them. As a moratorium is a unilateral measure, it is unlikely to include verification mechanisms and transparency obligations. The breach of a moratorium would be obvious if a country carried out a nuclear or an anti-satellite test, but what if the country was secretly building up its capabilities for such testing, with the intention of revoking its moratorium in future? In the case of fissile material, it is hard to know whether countries that have declared moratoria have stopped production.
Moratoria can also simply be ignored. They are only binding on the country that has announced them. In the current challenging geopolitical context, where trust between the major actors is low, the effectiveness of moratoria is lessened because adversaries do not believe moratoria will be honoured.
The current context also makes using moratoria as a confidence-building measure and first step towards negotiations difficult. Adversaries are reluctant to support each other’s initiatives for political reasons, even if they have no objection in principle to the idea. An example is the US DA-ASAT initiative — no country is opposed to it in principle, but some do not want to support an American initiative.
Moratoria can play an important norm-creating function. This is particularly true for nuclear testing. Given that the CTBT has still not entered into force and Russia recently de-ratified it, the moratorium given by the P5 is even more important. Although it could be revoked at any time, the political consequences of doing so are much higher now, as refraining from testing is considered an international norm.
Another problem with moratoria is what to do when they have not had the desired effect and doubts creep in about their usefulness. As they can be revoked at any time, thought should be given to restating them. In the case of nuclear testing, many feel that the P5 should renew their moratorium, ideally through a joint statement, with an additional commitment not to be the first to restart testing.
In the case of fissile material, the newly established Friends of FMCT group has recognised the value of voluntary moratoria as an important interim step pending the negotiation of a treaty. Thus far though, this recognition has not been effective in persuading China, India, and Pakistan to declare their own moratoria.
How Could Moratoria Be Used to Promote Arms Control Now?
How to use AI in the military domain responsibly is a hot topic in arms control and disarmament. However, the prospects for agreeing on new legally binding instruments in this area are not good.
States have been discussing lethal autonomous weapons systems (LAWS) for over a decade and they are still far apart on whether a new legally binding instrument is needed. In addition, regulating the use of AI in the military domain does not fit into a traditional arms control and disarmament agreement model of limiting or banning a particular category of weapons: AI is an enabler, not a weapon.
Can moratoria deal with some of the concerns around the use of AI in the military domain? Moratoria could be used to establish some broad principles and bring some of the reassurance that many are seeking, particularly around maintaining human control over decision-making by AI-based systems.
One area that many are worried about is the AI-nuclear nexus. AI is increasingly being used to support nuclear command, control, and communication (NC3). There are real concerns about the effect this will have on strategic stability, the risks posed by cyberattacks, and the consequences for leaders’ decision-making in times of crisis.
France, the United Kingdom, and the United States stated in a joint Working Paper to the 2022 Review Conference of the Nuclear Non-Proliferation Treaty that “consistent with long-standing policy, we will maintain human control and involvement for all actions critical to informing and executing sovereign decisions concerning nuclear weapons employment.” Although this statement was not in a stand-alone document, it is in the form of a moratorium and should be seen as politically binding on the three countries.
As noted earlier, at their 16 November meeting, Biden and Xi again affirmed the need to maintain human control over the decision to use nuclear weapons. If the P5 could make such a statement collectively, it would have more weight. I argued in a previous RSIS paper that such a joint statement would have even more weight if endorsed through a UN Security Council resolution.
Should the latest mandate of the Group of Governmental Experts on LAWS fail to produce agreement, then there would also be a case for moratoria on developing weapons systems that cannot comply with international humanitarian law. One advantage of this route would be that moratoria could be done nationally, leaving states free to choose their own formulation of words, thus getting round the endless debates over definitions. The disadvantage would be that differently worded moratoria would likely weaken their impact and lead to arguments around interpretation.
Conclusion
Moratoria have a mixed record and will never be a perfect solution. They have played a role as interim steps towards past agreements such as the CTBT; they have also failed to produce the desired agreements, such as an FMCT. It remains to be seen whether the US moratorium on DA-ASAT will lead to an agreement. It is significant that the United States chose this as an interim step, as it shows that moratoria are still part of the disarmament toolkit.
As states grapple with how to govern and regulate AI in the military domain, they should give serious consideration to moratoria. Done nationally, or even regionally, they could have an impactful norm-setting effect and would be a real achievement in today’s challenging and complex arms control and disarmament world.
Simon Cleobury is Head, Arms Control and Disarmament at the Geneva Centre for Security Policy. He was a Visiting Senior Fellow with the Military Transformations Programme at the S. Rajaratnam School of International Studies from October to November 2024.
SYNOPSIS
At a time when the prospects of concluding new arms control and disarmament agreements are not good, it is important to consider other tools that could help. One such tool is a moratorium — a state’s voluntary suspension of specific military activities. While there were some limits to the effectiveness of moratoria in the past, they may still be useful in the current context, where there is a lack of agreement on topics such as the use of AI in the military domain.
COMMENTARY
On 16 November 2024, presidents Joe Biden and Xi Jinping held a bilateral meeting in Lima. The two leaders agreed on the need for prudence and a sense of responsibility in developing AI technology in the military field. They also affirmed the need to maintain human control over the decision to use nuclear weapons. Although made in the context of a bilateral meeting, such an affirmation is a form of moratorium. Many want to see all nuclear weapons possessors make such a declaration.
On 21 November, Russia launched an intermediate-range ballistic missile on the Dnipro region in Ukraine. One reason for the strike was to respond to Ukraine’s long-range missile strikes on Russia, only days after their use had been authorised by the United Kingdom and United States. Another reason was Washington’s withdrawal from the Intermediate-Range Nuclear Forces (INF) Treaty and its subsequent refusal to agree to Moscow’s proposal to declare a moratorium on the development of missiles previously banned under the treaty. Had such a moratorium been in place, the Dnipro ballistic missile strike may not have happened. Russia has also recently stated that it would revoke this moratorium if the United States went ahead with plans to deploy short- and medium-range missiles to Europe and Asia.
What Is a Moratorium and Why Are Moratoria Used?
A moratorium, in the context of arms control and disarmament, is a voluntary suspension by a state of specific military activities, such as the testing, production, or deployment of weapons systems. Moratoria are politically binding on the countries that issue them. They are not legally binding, but most states, once politically bound, consider moratoria to have the same effect as a legal obligation.
States can use moratoria proactively to signal that they have decided to stop doing a certain thing and put pressure on others to make a similar commitment, which could then be turned into a legally binding agreement. A long-standing example in the space domain is the Russian initiative on “No First Placement of Weapons in Outer Space”. A more recent example is the US moratorium on direct-ascent anti-satellite testing (DA-ASAT).
A moratorium can also be an important political signal to create the environment for negotiating new legally binding instruments. In such cases, it demonstrates a state’s willingness to enter negotiations in good faith. An example is the moratorium announced by the “Permanent Five” UN Security Council members (P5) on nuclear testing, which preceded negotiations for the Comprehensive Nuclear Test Ban Treaty (CTBT).
Even when starting negotiations is unlikely, moratoria can still be useful as confidence-building measures. For example, France, Russia, the United Kingdom, and the United States have moratoria on the production of fissile material for use in nuclear weapons. However, their commitments have failed to persuade China and Pakistan, which do not have moratoria, to enter negotiations for a Fissile Material Cut-off Treaty (FMCT). India, which is in favour of an FMCT, does not see the value of a moratorium in persuading these two to come to the table.
China has adopted a similar approach with its policy of “No First Use of Nuclear Weapons”. Its undertaking that it will not be the first to use nuclear weapons is a form of moratorium and is intended as a first step towards negotiating its “Draft Treaty on No-First Use of Nuclear Weapons”.
Moratoria have also been suggested as a means of retaining some elements of a legally binding instrument that has fallen away. After Washington withdrew from the INF Treaty, citing Russian violations, Moscow proposed a moratorium on the development of missiles previously banned by the treaty.
Are Moratoria Effective?
Moratoria are voluntary — they can be revoked at any time. For some, this reality means moratoria have little value. On the contrary, the longer a moratorium is in place, the longer it has been adhered to, meaning it does have value. A good example here is the moratorium on nuclear testing.
However, that leads to the next criticism of moratoria, namely, that they are not verifiable and therefore there is no guarantee that countries which have issued them are abiding by them. As a moratorium is a unilateral measure, it is unlikely to include verification mechanisms and transparency obligations. The breach of a moratorium would be obvious if a country carried out a nuclear or an anti-satellite test, but what if the country was secretly building up its capabilities for such testing, with the intention of revoking its moratorium in future? In the case of fissile material, it is hard to know whether countries that have declared moratoria have stopped production.
Moratoria can also simply be ignored. They are only binding on the country that has announced them. In the current challenging geopolitical context, where trust between the major actors is low, the effectiveness of moratoria is lessened because adversaries do not believe moratoria will be honoured.
The current context also makes using moratoria as a confidence-building measure and first step towards negotiations difficult. Adversaries are reluctant to support each other’s initiatives for political reasons, even if they have no objection in principle to the idea. An example is the US DA-ASAT initiative — no country is opposed to it in principle, but some do not want to support an American initiative.
Moratoria can play an important norm-creating function. This is particularly true for nuclear testing. Given that the CTBT has still not entered into force and Russia recently de-ratified it, the moratorium given by the P5 is even more important. Although it could be revoked at any time, the political consequences of doing so are much higher now, as refraining from testing is considered an international norm.
Another problem with moratoria is what to do when they have not had the desired effect and doubts creep in about their usefulness. As they can be revoked at any time, thought should be given to restating them. In the case of nuclear testing, many feel that the P5 should renew their moratorium, ideally through a joint statement, with an additional commitment not to be the first to restart testing.
In the case of fissile material, the newly established Friends of FMCT group has recognised the value of voluntary moratoria as an important interim step pending the negotiation of a treaty. Thus far though, this recognition has not been effective in persuading China, India, and Pakistan to declare their own moratoria.
How Could Moratoria Be Used to Promote Arms Control Now?
How to use AI in the military domain responsibly is a hot topic in arms control and disarmament. However, the prospects for agreeing on new legally binding instruments in this area are not good.
States have been discussing lethal autonomous weapons systems (LAWS) for over a decade and they are still far apart on whether a new legally binding instrument is needed. In addition, regulating the use of AI in the military domain does not fit into a traditional arms control and disarmament agreement model of limiting or banning a particular category of weapons: AI is an enabler, not a weapon.
Can moratoria deal with some of the concerns around the use of AI in the military domain? Moratoria could be used to establish some broad principles and bring some of the reassurance that many are seeking, particularly around maintaining human control over decision-making by AI-based systems.
One area that many are worried about is the AI-nuclear nexus. AI is increasingly being used to support nuclear command, control, and communication (NC3). There are real concerns about the effect this will have on strategic stability, the risks posed by cyberattacks, and the consequences for leaders’ decision-making in times of crisis.
France, the United Kingdom, and the United States stated in a joint Working Paper to the 2022 Review Conference of the Nuclear Non-Proliferation Treaty that “consistent with long-standing policy, we will maintain human control and involvement for all actions critical to informing and executing sovereign decisions concerning nuclear weapons employment.” Although this statement was not in a stand-alone document, it is in the form of a moratorium and should be seen as politically binding on the three countries.
As noted earlier, at their 16 November meeting, Biden and Xi again affirmed the need to maintain human control over the decision to use nuclear weapons. If the P5 could make such a statement collectively, it would have more weight. I argued in a previous RSIS paper that such a joint statement would have even more weight if endorsed through a UN Security Council resolution.
Should the latest mandate of the Group of Governmental Experts on LAWS fail to produce agreement, then there would also be a case for moratoria on developing weapons systems that cannot comply with international humanitarian law. One advantage of this route would be that moratoria could be done nationally, leaving states free to choose their own formulation of words, thus getting round the endless debates over definitions. The disadvantage would be that differently worded moratoria would likely weaken their impact and lead to arguments around interpretation.
Conclusion
Moratoria have a mixed record and will never be a perfect solution. They have played a role as interim steps towards past agreements such as the CTBT; they have also failed to produce the desired agreements, such as an FMCT. It remains to be seen whether the US moratorium on DA-ASAT will lead to an agreement. It is significant that the United States chose this as an interim step, as it shows that moratoria are still part of the disarmament toolkit.
As states grapple with how to govern and regulate AI in the military domain, they should give serious consideration to moratoria. Done nationally, or even regionally, they could have an impactful norm-setting effect and would be a real achievement in today’s challenging and complex arms control and disarmament world.
Simon Cleobury is Head, Arms Control and Disarmament at the Geneva Centre for Security Policy. He was a Visiting Senior Fellow with the Military Transformations Programme at the S. Rajaratnam School of International Studies from October to November 2024.
