Dansk Interparlamentarisk Gruppes bestyrelse 2023-24
IPU Alm.del Bilag 6
Offentligt
2852452_0001.png
148th IPU Assembly
Geneva, 23–27 March 2024
Addressing the social and humanitarian impact
of autonomous weapon systems
and artificial intelligence
Resolution adopted by consensus
*
by the 148th IPU Assembly
(Geneva, 27 March 2024)
The 148th Assembly of the Inter-Parliamentary Union,
Acknowledging
that, while the applications of emerging technologies open up opportunities
for the development of humanity, they may also pose significant challenges to peace and international
security and may raise new questions about the role of humans in warfare, that regulation of autonomy in
the context of weapon systems requires a holistic understanding of its effects, and that human
decision-making and control must take account of all ethical, legal, humanitarian and security
implications,
Affirming
that any discussion on autonomous weapon systems (AWS) is subject to
international law, particularly the Charter of the United Nations and international humanitarian law (IHL),
Noting
the lack of an agreed definition of autonomous weapon systems, and
recalling
the
proposal of the International Committee of the Red Cross (ICRC) in which the term “autonomous
weapon systems” encompasses any weapon system with autonomy in its critical functions, which
means that it can select (i.e. search for, detect, identify or track) and attack (i.e. use force against,
neutralize, damage or destroy) targets without human intervention,
Recognizing
that the preservation of human control and judgment stands as a critical factor
in ensuring legal compliance and effectively addressing the ethical concerns that arise from the
deployment of AWS,
Gravely concerned
that AWS that have full autonomy in their critical functions could be able
to select and attack targets without human intervention,
Concerned
that the lack of control and explicit regulation at the international level around the
use of AWS could allow operators to violate international law, in particular the Charter of the United
Nations and IHL, without accountability, potentially infringing on the fundamental human rights enshrined
in national, regional and international legal frameworks, due to the absence of human judgement and
supervision and the lack of opportunities for timely intervention or mechanisms for deactivation over the
use of force,
Deeply concerned
about the possible negative consequences and impact of AWS on global
security and regional and international stability, including the risk of an emerging arms race, lowering the
threshold for conflict and proliferation, including to non-State actors, as outlined in United Nations
General Assembly resolution 78/241 on lethal autonomous weapon systems,
Concerned
that advancements in sophisticated military technologies, including artificial
intelligence (AI) and algorithmic data processing, may increase the risk of a new arms race, lowering the
threshold for conflict and proliferation, including to non-State actors, and putting peace and international
security at even greater risk, but
acknowledging
that a ban on research would be unrealistic, not least
when so much of the research in this field is conducted by both military and civilian players and AI still
has a very important role to play in civilian life,
#IPU148
IPU, Alm.del - 2023-24 - Bilag 6: Outcomes of the 148th Assembly in Geneva
-2-
Alarmed
by the possibility that AWS have the potential to become future weapons of
mass destruction as they combine two properties unique to such weapons: mass harm and lack of
human control to ensure they do not injure civilians,
Conscious
that human rights instruments guarantee the right to life, dignity and
integrity of persons,
Deeply concerned
that AWS could be used by armed groups and other non-State
actors to undermine national, regional and global security, causing profound social and
humanitarian impact,
Alarmed
by the evidence that individual recognition algorithms, including facial
recognition and automated decision algorithms have native embedded bias that is already
propagating gender and race discrimination and perpetrating injustices against socio-economically
disadvantaged people, the vulnerable and people with disabilities, and that AWS could be
deliberately programmed to target people bearing certain “markers” or identities including race,
gender or patterns of behaviour, and to apply force without human intervention, potentially leading
to disproportionate harm to specific groups, locations or communities,
Recalling,
without prejudice to Article 51 of the Charter of the United Nations, the
fundamental rule under IHL according to which the right of Parties to a conflict to choose their
means and methods of warfare is not unlimited, as stated in Article 35 (1) of the Additional
Protocol I (AP I) to the Geneva Conventions of 1949 relating to the protection of victims in
international armed conflicts, and as provided for by customary international law, as well as the
obligation stated in Article 36 of AP I, which requires States to conduct reviews of the study,
development, acquisition or adoption of all new weapons, means and methods of warfare in order to
determine whether their use is prohibited by IHL or any other rule of applicable international law,
Mindful
that, for decades, the international community has been actively keeping track
of the emerging issues of AWS, marked by key milestones in AWS regulatory governance,
including the report of the United Nations Special Rapporteur on Extrajudicial, Summary or
Arbitrary Executions in 2010, which brought the issue of lethal autonomous robotics and the
protection of life to the international spotlight and that, since 2013, the High Contracting Parties to
the Convention on Prohibitions or Restrictions on the Use of Certain Conventional Weapons Which
May Be Deemed to be Excessively Injurious or to Have Indiscriminate Effects (CCW), a key IHL
instrument, have been holding discussions on the issue and, in 2016, established an open-ended
Group of Governmental Experts (GGE) on emerging technologies in the area of lethal autonomous
weapon systems (LAWS),
Recognizing
the GGE as a key international forum at which a framework around
LAWS is being developed,
Noting
the fact that, during its 2023 meetings, the GGE underscored the need to
anticipate technological advancements in LAWS, urging strict adherence to IHL throughout the life
cycle of these systems, highlighting the need for limitations on targets and operational parameters,
coupled with appropriate training and instructions for human operators, and firmly stating that any
LAWS-based system unable to comply with international law should not be deployed,
Acknowledging
the adoption of resolution 78/241 by the United Nations General
Assembly in December 2023, which, inter alia, requests that the Secretary-General of the United
Nations submit a substantive report on the subject reflecting the full range of views received from
Member and observer States on ways to address the related challenges and concerns such
systems raise from humanitarian, legal, security, technological and ethical perspectives, and on the
role of humans in the use of force, and invite the views of international and regional organizations,
the ICRC, civil society, the scientific community and industry,
Recognizing
that, since 2018, the Secretary-General of the United Nations has
consistently maintained that AWS are politically unacceptable and morally repugnant and has
called for their prohibition under international law, and that when presenting his
New Agenda for
IPU, Alm.del - 2023-24 - Bilag 6: Outcomes of the 148th Assembly in Geneva
-3-
Peace
ahead of the Summit of the Future in 2024, he further called on States to adopt by 2026 a
legally binding instrument to prohibit AWS that function without human control or oversight and to
regulate all other types of AWS,
Recognizing also
that the United Nations Special Rapporteur on extrajudicial,
summary or arbitrary executions, the United Nations Special Rapporteur on the rights of persons
with disabilities, the United Nations Special Rapporteur on contemporary forms of racism, the
United Nations Special Rapporteur on counter-terrorism and human rights, the ICRC, civil society,
including through the Stop Killer Robots Campaign, the scientific community, and academia have
joined the call by the Secretary-General of the United Nations for a global prohibition on AWS,
Noting
that the landmark joint appeal made in 2023 by the United Nations
Secretary-General and the President of the ICRC underlined the urgency for States to negotiate a
new, binding international law on AWS to set clear prohibitions and restrictions on AWS by 2026,
Mindful
that many States and groups of countries have already been calling for the
establishment of a legally binding instrument to regulate, limit and/or prohibit the use of AWS, and,
at the same time,
cognisant
of the need for a unanimous decision, i.e. that all States agree to follow
the proposed limitations,
Recognizing
that urgent, concrete action is needed to develop international
approaches, particularly given that weapon systems with varying degrees of autonomy have
already been used in various conflicts,
Considering
the purposes and principles enshrined in the Charter of the United
Nations, international human rights law, and IHL and its founding principles of humanity, dictates of
public conscience and ethical perspectives,
Reaffirming
that, because IHL requires commanders and the users of weapons to be
able to anticipate and limit their effects, weapon systems must be predictable, and that the potential
“black box” effect arising from the integration of AI technologies could hinder compliance with these
obligations,
Gravely concerned
that the longer States wait to regulate AWS, the more likely the
flow and proliferation of such systems will continue on the market,
Stressing
the need to study an international regulatory framework for the use of AI to
regulate the harmful use of this technology,
Highlighting
that parliaments will have a significant role to play in raising awareness on
the social, humanitarian, legal and ethical implications of the use of AWS, and in supporting
governments with inputs to draft the text for an instrument to regulate such systems,
1.
Urges
parliaments and parliamentarians to actively and urgently engage in the debate
to address the threat to peace and security posed by AWS;
Strongly urges
parliaments to develop comprehensive national legislation to establish
regulatory frameworks governing the development, deployment and use of AWS, once
international agreement has been reached on a definition of “autonomous weapon
system” and on the distinction between full and partial autonomy as well as consensus
on the use and content of the term “meaningful human control”, taking
into account all
their ethical, legal, humanitarian and security implications and
including the prohibition
of AWS that function without human control or oversight, and which cannot be used in
compliance with IHL;
Calls on
parliaments to urge their governments to continue working through
international forums, including the United Nations and the GGE, on an instrument,
governance framework and regulations on autonomy in weapon systems, to ensure
compliance with international law, including IHL, and ethical perspectives, as well as
the prevention of the peace and security impact that autonomy in weapon systems
entails;
2.
3.
IPU, Alm.del - 2023-24 - Bilag 6: Outcomes of the 148th Assembly in Geneva
-4-
4.
Suggests
that parliaments encourage their governments to share their views with the
Secretary-General of the United Nations on ways to address challenges and concerns
raised by AWS in accordance with resolution 78/241 adopted by the United Nations
General Assembly in December 2023 and
A New Agenda for Peace,
which
recommends multilateral efforts for a legally binding instrument on LAWS and other
types of AWS by 2026;
Recommends
that parliaments and parliamentarians work with relevant stakeholders,
including the defence industry, civil society and academia, to understand, evaluate
and create safeguards in relation to both AI and AWS, including weapon system
designers, particularly regarding their compliance with existing law and with any
developments to the law that may occur in the future;
Encourages
parliaments to regularly assess and evaluate the application of new
technologies, to avoid these technologies creating a unilateral pressure on all citizens
that grants disproportionate powers to the parties when operating without proper
oversight, and to address the risks posed by facial recognition systems, including
hardware, software and algorithms, including to prevent gender and racial bias, that
may be integrated into AWS;
Urges
parliaments and parliamentarians to play a crucial role in holding governments
accountable regarding AWS, in ensuring quality in their governance, notably regarding
the imperative of retaining human control over the use of force, and transparency in
their design, development, operation, regulation and oversight, and in triggering
concrete action by governments and societies more broadly;
Calls on
parliaments to encourage governments to engage actively in the ongoing
discussions on LAWS at the GGE and to take all necessary efforts to support the
GGE’s work;
Also calls on
parliaments to strongly urge their governments to establish robust
frameworks for data protection to govern the development, deployment and use of
AWS, emphasizing the critical importance of safeguarding sensitive data and ensuring
ethical and responsible use of information;
Urges
parliaments to ensure the establishment of effective mechanisms to conduct
investigations, prosecution and punishment for violations of IHL arising from the use of
weapons with autonomous functionalities, thereby upholding individual responsibilities
and ensuring accountability for any breaches of ethical and legal standards;
Also urges
parliaments to request that their governments clearly define their own
responsibilities and those of the private sector and civil society with regard to AWS,
and adopt legislation that incorporates regulatory frameworks and safeguards to
ensure that such systems do not fall into criminal hands or into the hands of non-State
actors that operate outside the law, and that such laws are fully in line with
international human rights obligations;
Encourages
parliaments and parliamentarians to stimulate exchange of relevant good
practices between States, with due regard for national security regulations and
commercial restrictions on private information;
Recommends
that parliaments and parliamentarians: (a) allocate budgets to fund
plans, programmes, projects and actions to raise awareness of the need to prevent,
regulate, monitor and enforce human rights and safeguards related to AWS;
(b) advocate for the incorporation of comprehensive educational programmes on AI
and autonomous systems within national curricula at appropriate educational levels to
promote widespread understanding of both the potential benefits and the risks
associated with these technologies, including their ethical, legal, humanitarian and
security implications;
5.
6.
7.
8.
9.
10.
11.
12.
13.
IPU, Alm.del - 2023-24 - Bilag 6: Outcomes of the 148th Assembly in Geneva
2852452_0005.png
-5-
14.
Calls for
the adoption of measures to ensure the inclusion of a gender and
intersectional perspective based on United Nations Security Council resolution 1325
(2000), in discussions of AWS and military AI strategies;
Calls on
relevant parliamentary networks and IPU permanent observers to include
AWS on their agendas and to inform the IPU of their work and findings on the issue;
Invites
the IPU, through its relevant Standing Committee and specialized bodies, to keep
abreast of the issue and organize at the 151st Assembly a panel discussion, inviting
relevant parliamentary networks and IPU permanent observers to participate, aimed at
taking stock of the situation in advance of the 2026 deadline set by the
Secretary-General of the United Nations to adopt a legally binding instrument on AWS;
Suggests
that the IPU Secretariat compile and analyse a set of parliamentary good
practices and stocktaking related to the use of AI in the security and military sectors and
derived from the discussion within the IPU framework and other subsequent IPU
activities, including measures to remove bias in the algorithms that underpin AI systems
that are capable of autonomous analysis and actions;
Urges
the Secretary General of the IPU to share the present resolution and further
reports and publications related to AWS with the Secretary-General of the United
Nations for inclusion in the report mentioned in resolution 78/241 adopted by the
United Nations General Assembly in December 2023;
Invites
the IPU to regularly hold sessions for parliamentarians to discuss the latest
developments with AWS and AI and reassess their impacts in the military domain,
particularly with regard to specific concerns on human rights violations, meaningful
human control over the use of force and the ethical implications of these technologies;
Encourages
parliaments to implement their strategies to exercise more effective
parliamentary oversight functions and ensure that technological development, such as
AI, is only deployed to assist humans in certain tasks, without compromising
meaningful human control and intervention whenever it is needed.
15.
16.
17.
18.
19.
20.
*-
-
-
-
-
-
-
India, Iran (Islamic Republic of)
and the
Russian Federation
expressed their opposition to the entire
text of the resolution.
China
and
Lithuania
expressed a reservation on the entire text of the resolution.
Australia, New Zealand
and the
United Kingdom
expressed reservations on several paragraphs.
Cuba
and
France
expressed a reservation on the use of the term “autonomous weapon systems (AWS)”
without mentioning the characteristic of lethality (i.e. lethal autonomous weapon system – LAWS).
Türkiye
expressed reservations on preambular paragraphs 13, 17, 18, 20 and 21, and operative
paragraphs 2, 4 and 16.
Republic of Korea
expressed reservations on preambular paragraphs 18 and 19, and operative
paragraph 16.
Canada
expressed reservations on preambular paragraph 3 and operative paragraph 2 on the grounds
that they are too prescriptive for national governments.