# AI at a political crossroads: Mustafa Suleyman on the containment problem

Author: Fernando Nieto Lobato
Original publication: 2023-10-11
Spanish original: https://estrategiabyaleph.substack.com/p/estrategia-2-ia-la-encrucijada-politica
English URL: https://elcontemplador.github.io/estrategia-english/essays/002/
Status: Published translation

This is a translation of the original Spanish essay published on 11 October 2023. Its claims, examples and forecasts retain that historical context.

English publication: 2026-09-29

**Today, 11 October**, as this newsletter goes out, **the [Spanish edition](https://www.penguinlibros.com/es/ciencia-y-tecnologia/330715-libro-la-ola-que-viene-9788419399649) of *The Coming Wave* goes on sale**. This bestselling book on artificial intelligence is co-authored by Mustafa Suleyman, one of the world's leading AI experts and a co-founder of DeepMind, the company that currently forms the core of Google's artificial intelligence operations.

***The Coming Wave***, published in Spanish as ***[La ola que viene: Tecnología, poder y el gran dilema del siglo XXI](https://www.penguinlibros.com/es/ciencia-y-tecnologia/330715-libro-la-ola-que-viene-9788419399649)***, **offers a critical view of the implications of a range of developing technologies, led by AI, and raises significant questions about their impact on the global order**. Last week, as you may recall, we brought some of these reflections to this newsletter in our article on disinformation. According to Suleyman and his co-author Michael Bhaskar, **we stand on the threshold of a transformation that could propel humanity towards unprecedented technological and social progress. Yet this transition also carries the risk of disruption on a scale inconceivable in human history.**

Probably **the book's most significant contribution is its focus on the “containment problem”**. The authors warn of the emerging challenge of retaining control over increasingly powerful technologies, such as the convergence of AI and synthetic biology. These developments threaten to destabilise the “great bargain of the nation-state”, a pillar of the modern international order. Faced with this, Suleyman believes that **today's governments, generally ill-prepared for this new situation, face an existential problem**.

*Translation note: the quoted terms above are translated from the Spanish article; their precise wording in the English book has not been independently checked.*

## The proposed solution: ten steps towards containing advanced technologies

In the final part of the book, the authors propose **a ten-step roadmap for effectively containing these disruptive technologies**. These guidelines are a call to action which, although it may seem utopian, Suleyman and Bhaskar argue is essential to humanity's survival:

1. **Technical safety**: concrete technical measures to mitigate potential harm and maintain control.
2. **Audits**: a means of ensuring technological transparency and accountability.
3. **Chokepoints**: levers to slow development and give regulators and defensive technologies time.
4. **Makers**: ensuring that responsible developers build appropriate controls into technology from the outset.
5. **Businesses**: aligning the incentives of the organisations behind the technology with its containment.
6. **Government**: supporting governments so that they can build technology, regulate it and implement measures to mitigate harm.
7. **Alliances**: creating a system of international cooperation to harmonise laws and programmes.
8. **Culture**: a culture of sharing lessons and failures so that ways of addressing them can spread quickly.
9. **Movements**: all of this requires public participation at every level, including pressure to ensure that every component of these containment processes is fully accountable.
10. **Coherence**: ensuring that each element works in harmony with the others. Containment must be a virtuous circle of mutually reinforcing measures, rather than a cacophony of competing programmes riddled with gaps.

Given this newsletter's focus, **we will naturally examine steps 6 and 7 in a little more detail: these bear directly on political and governmental action concerning AI**.

[![Two robots debate above a semicircular parliamentary chamber, in soft watercolour-like colours.](https://elcontemplador.github.io/estrategia-english/assets/images/1f5d220b-5af8-48bb-9538-22a8a98d8a83_1024x1024.jpeg)](https://substackcdn.com/image/fetch/$s_!YLeM!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1f5d220b-5af8-48bb-9538-22a8a98d8a83_1024x1024.jpeg)

*Image generated with DALL·E 3 through ChatGPT. Prompt, translated from Spanish: An artistic illustration with watercolour brushstrokes for the newsletter “estrategIA”. Silhouettes of robots debating in a parliament symbolise AI's influence on politics, in soft, fluid colours.*

## The essential role of government

**In the sixth step, entitled “*Governments: survive, reform, regulate*”, the author sets out the urgent need for governments to take a more active and specialised role in technological regulation and development**. He argues that nation-states, with their control over fundamentals such as law and the economy, must address technology's emerging complexities in order to create and maintain resilient, secure social systems.

The central premise is that **technology**, although largely developed and operated by the private sector, **requires a deep understanding and comprehensive regulation on the part of government institutions**. A lack of technical competence within the public sector is not an option. The author maintains that **government investment in research and development is fundamental**. Such investment encourages a positive feedback loop in which government not only understands but also controls the implications and applications of emerging technology.

To achieve effective control, **the authors recommend that governments invest in talent and internal capabilities**, rather than relying solely on external consultants or suppliers. This approach will enable greater responsibility for, and ownership of, technological and political decisions. They also **advocate creating specialised government roles, such as a minister for emerging technology**, to oversee and regulate technological development effectively.

The government's role extends beyond regulation alone. **It must include detailed monitoring of how technology is used and how it affects the population.** This transparency will not only inform public policy but also educate the public about technology's risks and benefits.

On regulation, the book addresses the need to reassess licensing regimes for technological development. It proposes a more differentiated system based on the size and capabilities of the technological model in question. This would entail **paying particular attention to AI laboratories working on more fundamental and potentially dangerous capabilities**.

It also points to **the need for a complete overhaul of the tax system to fund safety and social welfare in the age of automation**. **It proposes taxing capital more heavily than labour and suggests initiatives such as a universal basic income** as possible solutions to the economic challenges accompanying technological development.

Finally, the book stresses that **no government can tackle these challenges in isolation**. **Technological regulation must be an international effort, particularly because technology knows no borders. This is the subject of step 7**, which addresses the need for international regulation and cooperation to manage technological advances, especially in areas such as artificial intelligence and biotechnology. The book cites historical examples of successful cooperation, such as the 1995 Protocol on Blinding Laser Weapons, to argue that international cooperation is possible even on sensitive issues. Suleyman suggests that **the most effective way to address the risks associated with emerging technologies is through strong, committed international alliances, rather than reliance on self-regulation or national rules**.

In this context, **the book discusses the need to create a global institution devoted to technology, similar to the International Atomic Energy Agency, which could act as an AI auditing authority**. This body would focus on global transparency and monitoring based on objective criteria, and would provide a foundation for government regulatory efforts.

Fernando Nieto Lobato

*Director of Digital Innovation, Institución Educativa ALEPH*