Recent events show the growing ability of artificial intelligence (AI) to compromise important systems. While the development of frontier AI remains concentrated in the United States and China, we argue that Canada and the European Union are well placed to lead the world’s middle powers on AI safety. Their upcoming summit in Montreal on 29-30 October is an opportunity to turn political commitments into practical action: a joint partnership for AI incident reporting and crisis preparedness that could become a model for international AI governance.
In both of these cases, the agents were still being trained, tested, or only internally deployed inside a US frontier AI company. No human directed them to do what they did. This pattern shows there is a need not only to defend against aggressive uses of AI by hostile actors, but also to establish dedicated systems to prevent and prepare for AI incidents even when there is no malicious intent or human threat actor involved.
The EU-Canada alliance has blossomed in recent years, including strong commitments to work together on AI. During her State of the Union address last month, European Commission President Ursula von der Leyen offered Canada a path to become the EU’s “first associate member” and urged cooperation on “early warning, AI security and much more.” Canadian Prime Minister Mark Carney reacted positively, saying the two powers can join forces on AI safety protocols, common standards and transparency.
Both leaders are also among the signatories of the “Call for Control” launched at the UN General Assembly, which, among other things, calls for “shared reporting of serious safety incidents”. However, expected deliverables from the upcoming EU-Canada summit reportedly do not contain much on AI safety.
In short, on AI safety, political will is on the record, while the needed mechanisms are still lacking. There is still opportunity and time to change that. To that end, we propose a bilateral EU-Canada AI Crisis Preparedness and Early Warning Partnership, built from existing parts.
Firstly, the EU and Canada should create a dedicated mechanism to share incident-relevant information with one another, so that both partners can better monitor, understand, respond to, prepare for, and prevent AI incidents and crises. This requires establishing trusted cross-border communication channels enabling the exchange of information on AI incidents and near-misses, agreeing upfront on what information to exchange, confidentiality rules and which authorities should have access to the information. Such a mechanism would enable both the EU and Canada to learn from what the other side sees, rather than operating in isolation.
This information-sharing mechanism could be included in an Administrative Arrangement between the EU AI Office and the Canadian AI Safety Institute, perhaps modeled on the 2024 EU-Singapore arrangement on AI safety. Such an arrangement could enable the EU and Canada to establish clear protocols for sharing serious incidents and near-misses (as allowed by the EU AI Act’s Article 78(5)).
Secondly, the EU and Canada should create a common reporting format for AI incidents.
Thirdly, systematic sharing of standardised information on AI incidents should become the foundation for establishing broader EU-Canada collaboration on AI safety. Ultimately, middle powers like EU Member States and Canada must not simply report AI incidents, but work together to prevent crises in the first place. Hence, information-sharing should directly feed into and support domestic incident response capacity, forensic analysis and regularly updated crisis prevention and preparedness measures. Incident reporting must generate pooled lessons and actionable insights, not simply a shared incident database.
The structures to organise this collaboration could be enabled by the proposed Administrative Arrangement on AI safety, but also by other existing mechanisms. AI should form a major part of the collaboration between EU-Canada crisis centres established by the 2025 Security and Defence Partnership. We suggest the centres already begin holding joint AI crisis prevention and preparedness-focused tabletop exercises as early as next year. The EU should also fund joint research on international AI incident detection, root-cause analysis, and crisis prevention protocols through the Horizon Europe programme, to which Canada is associated.
Crucially, building up EU-Canada information-sharing and collaboration on AI incidents does not require working from scratch. The EU already runs cooperation across 27 Member States through its formal cooperation network of national computer incident-response teams (CSIRTs Network) and its cyber-crisis network (EU-CyCLONe). Both networks facilitate routine and proactive cross-border sharing of relevant information, measures,best practices and frameworks for cyber threats, incidents,and vulnerabilities. Canada could, with eventual adaptations, become associated with these networks.
An EU-Canada partnership on AI safety would also give practical momentum to the two powers’ Digital Partnership, launched in 2023 with AI among its core priorities. Cooperation on incident reporting and crisis preparedness could build the trust and technical relationships needed for a wider AI partnership: joint research, shared evaluation capacities and collaboration on trustworthy AI applications in healthcare, industry and public services.
This proposed partnership could serve as a bilateral pilot for a model of transnational AI governance that could be scaled globally. A workable EU-Canada model may well prove to be appealing to other middle powers such as Australia, Japan, Singapore and the United Kingdom, based on a shared interest in maximising AI safety and security within a rules-based international order. It could also feed into the G20 and G7, where Carney has floated a “technology stability board”, and into the international network of AI safety institutes.
Middle powers are not doomed to being spectators before the AI superpowers. Canada and the EU do not need to dominate frontier AI development to help make the technology safe. In Montreal, their leaders should commit to a practical partnership: trusted channels for early warning, compatible incident reports and growing collaboration on AI as part of their wider digital partnership. These are achievable steps that would turn political declarations into a capacity to act, while offering other countries a model they could join or replicate. In this way, an EU-Canada partnership on AI safety could constitute a decisive step towards a safer world.
Across both Brussels and New Delhi, governing artificial intelligence has become a question of political economy as much as of technology policy. Choices about regulatory design, infrastructure access, and standards will determine which actors capture AI’s gains, which bear its risks, and which retain the capacity to govern it at all. Right now, those choices are being made largely within a framework defined by US and Chinese interests. Yet for all the energy spent in the US-China AI rivalry, one of the most promising partnerships in global AI governance is the one emerging between the European Union and India.
This article examines the foundations, opportunities, and challenges of EU-India AI cooperation. It argues that shared strategic vulnerabilities, complementary institutional strengths, and recent multilateral milestones have created a window for Brussels and New Delhi to deepen their collaboration and exert meaningful influence on the wider architecture of global AI governance.
The publication of Pope Leo XIV’s first encyclical, Magnifica Humanitas, marks an important moment not only for Catholic social doctrine, but also more broadly for a reflection on the foundations of political, social and international order. While the document is formally centred on the challenges posed by artificial intelligence, it in fact contains a much wider diagnosis of the current condition of humanity and of global coexistence—including a sharp analysis of the current crisis of international order.
Historically, major social encyclicals have not only clarified doctrinal principles, but have also shaped intellectual, political and social agendas well beyond ecclesial circles. One might think of Rerum Novarum and its long-lasting influence on Catholic approaches to socio-economic questions, or of Quadragesimo Anno and its central role in developing the principle of subsidiarity. It would therefore not be surprising if Magnifica Humanitas were to play a similar role in the coming years, especially with regard to questions of governance, technology and international order.
For Christian Democrats, conservatives and, more broadly, for all those political traditions open to the contribution of religious thought to public life, this document deserves careful and sustained reflection. Although the problem of international order is not presented as the central thematic focus of the encyclical, it is in fact one of its underlying concerns. The present contribution aims to explore precisely what kind of reflections Magnifica Humanitas may inspire on this topic.
Although barely reported in Europe, South Korea has just experienced a terrifying glimpse of how AI stocks increasing market dominance – coupled with modern trading technology and lax financial regulations – can cause extreme volatility and wealth destruction.
The South Korean stock exchange (KOSPI) doubled in value between January and June 2026 before plummeting by nearly 40% over the following month.
Although designed to represent the entire South Korean economy, the KOSPI had come to represent a one-way bet on the AI boom. Over 50% of the entire index was represented by just two firms – Samsung and SK Hynix. But this wasn’t a bet on TV’s and high-end electronics – no, this symbolised a wager on these companies’ critical roles in providing High Bandwidth Memory (HBM) chips and other data centre hardware to the AI juggernaut.
This is a classic example of over-concentration risk. And one which the United States would do well to heed. Already, AI-linked stocks account for around 45% of the S&P 500’s total market capitalisation. And this is before companies such as OpenAI and Anthropic even list on the US exchanges!
Even more worrying is that existing AI ‘Hyperscalers’ (technology companies that provide global networks of data centres) now dominate earnings and debt flows. Just three Hyperscalers —Alphabet, Amazon, and Meta — are responsible for roughly 70% of the S&P 500’s entire earnings growth expectations. Hyperscalers alone have borrowed over $220 billion so far in 2026 to finance their AI ambitions.
Yet, it’s not just concentration risk that should keep global policymakers awake at night.
The South Korean example also highlights how modern trading technology – the ability to trade instantly on your smartphone has pitfalls for millions of ordinary (and often not fully informed) investors. As noted by the Wall Street Journal, “more ordinary Korean investors joined the frenzy, including stay-at-home mothers, students and retirees cashing in their pensions. Many newcomers became stock-rich almost overnight. Margin loan balances rose by $7.9 billion to $27.1 billion in six months, as more investors tried to boost returns”.
Amplified by social media, local retail investors, known as “ants,” used online coordination and debt to pile into Exchange-Traded Funds (ETF) that were based not just on the investors’ initial investment, but also included a borrowed component. These ‘leveraged ETFs’ often use more complex financial instruments (such as derivatives) and debt to multiply the daily returns of an underlying index or asset. In reality, these products magnify both gains and losses – and thus provide a poor substitute for a sustainable, long-term investment strategy.
The Korean boom-and-bust cycle was also fuelled by direct political amplification. ABC News Australia described it succinctly: “the South Korean government contributed to the 2026 stock market crash by aggressively encouraging retail investment and loosening regulations on debt-financed trading products just as a speculative tech boom reached its peak”.
This lack of regulation was directly associated with South Korean President Lee Jae Myung’s desire to erase the traditional discount attributable to South Korean stocks compared to global peers. This so-called “Korea discount” has traditionally been based on corporate governance concerns concerning the complex ownership structures of the large conglomerates which dominate the South Korean economy.
Central to this strategy of increasing domestic stock market investment was the popularisation of leveraged ETFs. The results, as we have seen, have been a traumatic (and expensive) lesson for millions of South Koreans.
In a wider context, it’s clear that AI stocks (or stocks basing their future earnings on the continued stratospheric growth of the AI ecosystem) pose a fundamental risk to global macroeconomic stability. This risk is exacerbated by rapidly expanding public and private debt levels, such as in the United States.
South Korea also shows that this economic risk is magnified when combined with modern trading technology, lax regulation of available financial products, and limited financial literacy among the wider population.
Perhaps also, the volatility in Korea highlights the susceptibility of huge swathes of the population to investments promising immediate and incredible returns. This is at least partially attributable to the increasing disconnect felt in most developed societies between social mobility and earned income. For many lacking inherited family wealth, the roulette of poorly understood investments remains a potential pathway to the traditional middle-class dream of economic security.
Europe will not be immune to the contagion caused by interlinkage of AI, debt and often antiquated financial regulation. We would do well to heed the South Korean warning.
This policy brief argues that the General Data Protection Regulation (GDPR) should be refocused on the duty of loyalty in data relationships. This would allow a much-needed simplification of the Artificial Intelligence Act (AI Act). The purpose is to establish a regulatory focus on addressing the data power wielded by powerful entities online, while improving competitiveness. This can be achieved by the light-touch regulation of data processing entities that do not pose a significant risk to democracy and rights. The duty of loyalty requires that anybody processing personal data must act in the best interests of the people who may be affected by that processing. The higher the risk related to data processing, the greater the regulatory demands posed by the duty of loyalty. Such an approach would allow regulatory attention to focus on data power in the infosphere. Entities that hold such power should face strict duties and clear prohibitions; those that do not should meet only proportionate requirements.
The announcement of the provisional agreement to establish the first comprehensive regulatory framework on AI by the European Union marks a historic development. At a time when even the creators of technological innovation publicly express concerns about the potential for unchecked dominance of machines over humans, Europe is taking a leading initiative. It aims to impose fundamental principles and values on the technological development of the 4th Industrial Revolution.
The agreement aims to ensure that AI is utilised in a transparent, fair, safe, and environmentally friendly manner without limiting possibilities and opportunities for European AI-related start-ups. Among other provisions, it guarantees that AI-generated content will always be labelled as such, and AI systems interacting with humans will be required to inform the user that they are in contact with a machine. Human control systems for machines and the installation of risk management systems are also foreseen.
Simultaneously, a distinction is made between systems classified as ‘high-risk,’ such as those used in sensitive areas like critical infrastructure, training, human resources, and public order. Particularly in the latter category, the use of remote biometric identification systems in public places is restricted to avoid mass surveillance of populations. The text goes beyond theoretical considerations and establishes a necessary control mechanism—the European AI Office, which will coordinate compliance and enforcement, having the authority to impose significant financial penalties.
Despite reservations expressed about the potential impact on Europe’s technological competitiveness, a balanced approach has been taken to ensure that research is not restricted while implementing strict regulations, mainly on large-scale applications. In this framework, the technology market itself acknowledges that the new legislation leaves companies room for manoeuvre, despite the restrictions.
While Europe may not be in a position to take the technological lead in the development of the 4th Industrial Revolution, it has demonstrated its ability to use its organised market and institutions to regulate an environment which risks becoming chaotic.
Let us make no mistake: technology will always advance faster than bureaucratic negotiations of regulations.
Within four months, we transitioned from ChatGPT 3 to ChatGPT 4, representing a shift from a system with 175 billion parameters to one with a trillion parameters.
The time it took for ChatGPT to ‘reinvent’ itself was not enough to pass the regulation from the European Parliament to the Council, a distance of less than one kilometre in the heart of Brussels. Considering that the text was initially presented in 2021 and the regulation is not expected to take effect before 2025, the significant difference in the reaction reflexes becomes understandable. Nonetheless, the importance of the European decision remains substantial. The EU is the first international entity to successfully impose regulatory rules on the development of artificial intelligence, establishing crucial parameters of transparency, accountability, and control, while making the technology more human-centred.
Europe has proven itself to be the most sensitive international actor in protecting human rights and upholding fundamental principles and values. It is now laying the foundation for further development of a stronger international regulatory framework, providing a model that could be adopted more widely with country-specific adaptations. Additionally, the rules are likely to be applied outside the EU for platform functionality reasons.
It also paves the way for a similar intervention at the level of International Law, which would have been imperative had the EU not taken action, given that the UN recently seems to have sunk into a quagmire of inefficiency and mere observation of major international developments.
In any case, the EU appears to have learned a crucial lesson from the uncontrolled growth of technological giants. It has taken steps to address this issue belatedly, as these wholly unregulated platforms became available for the dissemination of fake news and hate speech, resulting in profound social and political consequences. That is why the EU’s regulatory intervention on AI, following the Digital Service Act and other related legislation, is a step of historic importance, no matter how modest it may appear today.
Artificial Intelligence (AI) is changing our world. This new phenomenon carries many threats, but also offers many opportunities. We need to find a suitable framework to support trustworthy AI. A key challenge remains: can we, as humans, retain control over the technology or will the technology take control of humanity? In responding to this challenge, the following question needs to be considered: What kinds of tools are needed, not only to keep control of AI development, but foremost to multiply the possible opportunities it offers?
The current pandemic has shown how useful and important AI can be in helping us to fight COVID-19. Moreover, it has clearly demonstrated that we cannot afford not to utilise it, nor do we have time to lose with regard to its development.
Hence, it is our responsibility to urgently establish an adequate framework for the development of AI systems based on a revision of the existing law and followed by possible new legislative proposals with a clear focus on future-proof tools. We have to generate a suitable governance model that not only has its foundation in law, but that also ensures democratic oversight through the open and collaborative participation of all partners and the validation of AI solutions by science and society. We should build trustworthy AI based on a human-centric and principled approach. The practical implementation of ethical rules in the design of AI (through the existing ex post model of analysing the consequences, including unintended ones, as well as a new ex ante model that provides an impact assessment in the early stages of development) and the evaluation of the everyday functioning of AI systems are essential.
It will not be possible to develop AI and claim all its economic and social benefits without a clear model for data use (including flows, collection and processing) that fully respects fundamental rights and the principles of cybersecurity. It will not be possible to build trustworthy AI without transparent rules for the relationships between its users (workers, citizens and consumers) and AI designers, developers and deployers (with the symmetry of information required, e.g. practical schemes for ‘explainability’). It will not be possible to accurately implement various AI functionalities without undertaking risk assessments and introducing mechanisms to manage those risks.
To achieve all of the above, we need compromises at various levels: between European institutions and stakeholders (businesses, citizens and consumers, taking into account rights), between European 10 institutions and member states (based on common and harmonised solutions), and between political groups, which are currently more focused on their differences than similarities. How can these compromises be achieved swiftly?
The answer is multidimensional and complex; however, we should be brave enough to pursue it. Paradoxically, the unfortunate experience of COVID-19 has brought a lot of positive momentum to our search for answers, proving to be a real AI development game-changer.
Artificial Intelligence (AI) has emerged as the main engine of growth of the Fourth Industrial Revolution, owing to its naturally cross-cutting, general-purpose nature. From a military perspective, the range of potential applications is at least as vast as the current range of tasks that require human cognition, e.g., analysing and classifying visual data, organising logistics, operating vehicles, or tracking and engaging hostile targets. How can Western nations – by which I mean those nations that are members of either NATO or the European Union (or both) – make the most out of the rise of AI, bearing in mind its potential defence applications?
To provide the best experiences, we use technologies like cookies to store and/or access device information.
Functional
Always active
The technical storage or access is strictly necessary for the legitimate purpose of enabling the use of a specific service explicitly requested by the subscriber or user, or for the sole purpose of carrying out the transmission of a communication over an electronic communications network.
Preferences
The technical storage or access is necessary for the legitimate purpose of storing preferences that are not requested by the subscriber or user.
Statistics
The technical storage or access that is used exclusively for statistical purposes.The technical storage or access that is used exclusively for anonymous statistical purposes. Without a subpoena, voluntary compliance on the part of your Internet Service Provider, or additional records from a third party, information stored or retrieved for this purpose alone cannot usually be used to identify you.
Marketing
The technical storage or access is required to create user profiles to send advertising, or to track the user on a website or across several websites for similar marketing purposes.
Loading...
