The Task Force On Cooperation in Greater Europe: Istanbul, 8-9 April 2019

1. New military technology is inevitable and can be stabilising.

2. But destabilising military technologies are appearing and proliferating at speed.

3. This is serious. It is a very difficult systems problem.  It carries rising risks of misjudgement, miscalculation, misinterpretation, mischief and accident.

4. And it is urgent.  The more delay, the more danger and the harder the introduction of regulatory measures.

5. It demands attention by states, experts, civil society and industry. The combination of multi-tech, multi-actor, complex, fast-paced, poorly understood, cross-domain effects is a phenomenon and a problem in its own right, independent of the dangers in any given technology domain.

6. The goal(s) should be:

• Strategic stability
• Minimising harms

7. These goals cannot be achieved without:

• Prioritarisation
• New paradigms for arms control

8.  Prioritisation. We suggest the following criteria for action:

• Prioritise the pursuit of existential common interests between states
• Prioritise tackling those technologies that most destabilise the nuclear order.Examples include the entanglement of nuclear systems with other military capabilities (as in satellites carrying nuclear and conventional command and control); the blurring of boundaries between the use of nuclear and other military technologies; the loss of clarity in nuclear deterrence models; above all, the threat to nuclear stability from cyber.
• Prioritise tackling the clearest pathways to nuclear danger.Identify worst case outcomes and work back.
• Prioritise international early wins: the universal application of international law to new military technologies; closing off emerging technology danger areas (ideally, both Lethal Autonomous Weapons Systems (LAWS) and military Artificial Intelligence); tackling linkages to existing problems (the crisis of arms control; the pollution of outer space; erosion of the 20th century nuclear order)

9.  New paradigms for arms control.

• This need is real and urgent.Progress is thin. States, experts, civil society and industry all must address where more research is need. States need to invest in such research.
• This does not at all diminish the need to keep working and negotiating on each type of destabilising military technology.For example:
  • Cyber: clarify terminology.Consider chemical weapons-style arms control models
  • Space: International frameworks are needed to regulate dual use capabilities.
  • LAWS: Agree the place and role of the human in conflict.
• It does not at all diminish the need to protect and build upon existing regimes: US-Russia strategic nuclear control; the Nuclear Non-Proliferation Treaty; the 5 outer space treaties.
• But old, quantitative, static, comprehensive, often bilateral arms control is less and less likely to work. There is still a nuclear order but it is eroding in the face of multi-tech, multi-player, entanglement.The conduct of conflict is changing fast.
• Future regimes are likely to be qualitative more than quantitative, more about deterrence by denial and by resilience than deterrence by punishment, about behaviours and attribution rather than constraints on capabilities, about creating norms more than treaties, about international social coalitions not just states, about economics not just security, about transparency and intentions, about risk reduction and global change management – security ecologies rather than architectures:
  • The civil sector drives new tech, no longer the military;
  • Industry has massive stakes and more power than many states but avoids responsibility;
  • Industry may introduce some measures of self-restraint and may develop “ethical guidelines” for their employees (researchers);
  • Civil society will increasingly hold industry to account as they are doing with the technology companies (cf. Microsoft and Google employees dictating terms on their companies’ military contracts);
  • States are losing control.And they cannot do it all.They will be increasingly vulnerable under ’multipolarity without multilateralism’.So much so that they may have a shared interest in collaboration to preserve state control and state authority.

10. Decisionmakers might consider, for example:

• Codes of Conduct, Rules of the Road, Responsible Behaviours, Expectations clearly communicated through good declaratory policy
• Criminalisation of unacceptable behaviours
• National action (e.g. legislation) pending international action
• Transparency and trust building
• Both commonalities across tech domains and linkages between them, seeking cross-domain deals on managing entanglement of different military technologies, especially nuclear.

11. Future regulatory regimes need:

• Heightened awareness, political will, appropriate skill sets among decision makers, advocacy by stakeholders
• Conceptual clarity, new taxonomies, terminological precision, honest public discourse: ‘arms races’ are never really ‘won’.Perhaps talk not of ‘arms control’ but of ‘tech management’
• Breaking down the expert and conceptual silos to promote more fertile thinking and collaboration.Sharing of best practice across tech domains. (For example, all space activity must be attributed to a state. This is not (yet) the case in the cyber domain.)
• Engagement of a wide range of stakeholders that have a ‘responsibility to protect’.

12.  States have not yet seriously confronted these challenges.  Instead, states that might be expected to carry particular responsibilities have dealt blow after blow to global arms control and disarmament diplomacy.

Global Relations Forum (GIF/GRF)

European Leadership Network (ELN)

Russian International Affairs Council (RIAC)

Istituto Affari Internazionali (IAI)

May 2019