Excellencies,
ladies and gentlemen,
dear friends and guests,
I am excited that we managed to compete with the EURO 2024 tournament today in drawing such an impressive crowd to Berlin with a less joyful topic than football.
But the subject we will be discussing today is also on everybody’s mind these days.
You may have seen the current The Economist title “War and AI”
The possibilities of artificial intelligence are omnipresent in our media, and companies at the forefront of this technology are seeing their share prices skyrocketing. Although we have already been witnesses of enormous progress in computer science, AI is undoubtedly a transformative technology that will profoundly change our world in many ways. AI is a cross-cutting issue, and needs to be tackled as such. In response, earlier this month the German Foreign Office adopted an AI Charter with principles for a rules and values-based application of AI.
This is work that will keep us busy for a long time.
As with every technology developed by humans, AI offers great opportunities, but also risks. There are even fears we may cross a line and empower machines to an extent that they might spin out of control. Every new technology can be used not only for civilian, but also for military purposes. AI is no exception.
AI is already a factor in the shaping of future conventional warfare, allowing weapons systems to process vast amounts of data in complex and dynamic scenarios and assisting human decision making. The development of lethal autonomous weapons systems raises complex questions, including legal and ethical ones, and is subject of an intense debate. Germany is strongly and actively engaged in the multilateral discussions of these questions, as well as in international processes aiming to ensure responsible use of AI in the military domain.
The questions at hand get still another dimension when it comes to the use of AI in the design or the command and control of weapons of mass destruction. Hollywood of course knew the answer already years ago with its “Terminator” blockbuster where an AI system called “Skynet” designed to run nuclear command systems turns against its human masters.
While this is and hopefully will stay science fiction, AI will have profound implications also for WMD arms control. We are just at the beginning of understanding what these implications are, and we need to dive deeper in our assessment of the risks and opportunities.
Media headlines, not surprisingly, stress the dangers, and in rather drastic terms: We hear of a “killer AI” which “invented 40,000 lethal chemical weapons in just 6 hours”. We are warned that AI may give terrorists easy access to information for building a chemical or biological weapon. We are told AI based nuclear command and control systems may “hallucinate” and accidentally unleash nuclear
Armageddon. What we need is get beyond the hype of these doomsday stories and towards a realistic understanding.
We should start with a sober look on what AI actually is – a technology, albeit a very powerful one, designed and used by humans. It is another tool, enabling humans to do things faster or more efficiently than before, or doing new things that were impossible before. AI is a highly powerful enabler. We see this in the convergence of AI and other emerging technologies. As we will see later today, using AI as a tool to design new or altered synthetic pathogens could pose a serious challenge to the ban on biological weapons. AI could make existing tools to undermine our security a lot more dangerous, including the use of AI for disinformation. It is difficult to predict whether AI increases or reduces vulnerabilities to cybersecurity threats, but it’s a key question if you consider using AI in command and control systems for nuclear weapons.
With this conference, we want to shed more light on these questions. Today’s event continues our conference line on capturing new technology and rethinking arms control in response. We want to identify the most relevant challenges coming up. We want to start the debate on possible responses, how they could look like, and how we should organise to develop them.
Tackling those challenges is a team effort. We all need to work together! I am glad that work has started. I recall the AI Safety Summit organized by the British government last November which looked at the dangers coming from rogue non-state actors. France will carry the effort forward with an AI Action Summit next year. I would also like to mention the US initiative for a Political Declaration on Responsible Military Use of AI and Autonomy. We need to broaden these efforts. We must examine what steps we can take within the multilateral arms control frameworks. We need to look at how export control regimes need to be adapted and where we need to readjust the line between freedom of scientific research and scientific responsibility. While doing this, we should neither hamper scientific progress to be used to the benefit of humankind, nor should we create undue burdens for a burgeoning industry in a global competition.
In today’s conference we will start with a closer look into the impact on chemical weapons. It is my pleasure to welcome the Director General of the Organisation for the Prohibition of Chemical Weapons, Fernando Arias, as our keynote speaker today. Fernando, we are honoured that you accepted our invitation.
It is not a coincidence that we don’t follow the alphabetic order of weapons of mass destruction, but start with a “C”. You have been pointing at the challenges posed by AI for some time already, including in your important address to last year’s CWC Review Conference. Our joint objective is to ensure that technological developments do not endanger the global ban of chemical weapons. I am convinced the keynote we are about to hear from Director General Arias will be an excellent start to our discussion today.
Before handing over, let me encourage everybody to engage in an open discussion today. To facilitate this, we decided to place this event under Chatham House rules, unless otherwise indicated. I thank you all for coming to Berlin today, to the Auswärtiges Amt. I am grateful for your attention, and I am very pleased to hand over to Director General Arias. Thank you.