Artificial intelligence (AI) may still hold the potential to solve some of the world’s most intractable problems and help fulfill the UN Sustainable Development Goals (SDGs), but when it comes to risks to privacy and civil liberties, AI already has been a game changer in favor of authoritarian states. AI-enabled tools have turbocharged every pre-existing form of repression including: mass and targeted surveillance, censorship, and the spread of propaganda. Contrary to the original expectation that it would be impossible for repressive states to control the open internet, AI has facilitated a whole new level of state control over communications infrastructure and the information realm. Its technological advantages include scaled capacity to scan for forbidden content and filter out dissenting views. In the other direction — the production of ideas — autocrats have found new ability to control public narratives and shape civic discourse with AI-generated and amplified content. New social engineering tools, such as China’s social-credit system, mold citizens’ motivations and behaviors. Beyond violating privacy and civil liberties, these systems have the potential to destroy, in significant part, human agency and human dignity.

The larger threat posed by all these AI-enabled technologies is that they are facilitating the spread of digital authoritarianism: an encompassing techno-social system and governance model that involves control and security for the state as opposed to liberty and security for citizens. Rather than view the challenge as a series of discrete apps used for repression, democracies should see digital authoritarianism through the lens of system rivalry and recognize that they face competition from a powerful, repressive governance model spreading around the world.

This model is being propagated through a variety of means. It certainly includes the diffusion of technology, but it also includes the diffusion of values, norms, and concepts related to appropriate uses of and constraints on technology. The authoritarian model is also spread through propaganda and economic coercion, sometime referred to as “sharp power,” and even through concerted efforts to influence tech standard-setting bodies, where repressive potential can be embedded in tech protocols for the future. These elements generally come as a package deal.

China as Model of Digital Authoritarianism

The starkest example of digital authoritarianism is China’s version, which is manifesting global influence on multiple layers. First, China has become a role-model of AI-driven domestic repression, with highly escalated surveillance capacities in their own “smart cities” and panopticon-level control in regions deemed “security threats,” such as Xinjiang. They have demonstrated stunningly effective control over the domestic information realm and the ability to steers citizens’ behavior with powerful incentives built into their social credit system. China also is advancing development of a sovereign digital currency that will substantially enhance their repressive powers at home.

Second, China is exporting these repressive capacities and normalizing their use around the world. Through broader economic trade and development initiatives, such as its Belt and Road and Digital Silk Road, China has built entire infrastructure systems through which it has gained leverage over fragile states that will last for decades to come, as well as new sources of data that can be sucked back to Beijing.

Third, China is shaping debate in international normative arenas by flooding the zone of multilateral tech-related diplomacy. Their advocacy has been very effective within normative bodies like the UN Human Rights Council, where China has swayed the majority of delegations to support their repressive uses of technology in Hong Kong and Xinjiang. They also have demonstrated an ability to exert influence at tech standard-setting bodies, like the International Telecommunications Industry (ITU), where interoperability standards for the future are set. Their aim has been to push China’s preferred protocols as the global default for Internet of things (IOT) and other emerging technologies.

Fourth, in the global marketplace of ideas, China is spreading propaganda about the weakness of democracy, using so-called “Wolf Warrior” diplomats who aggressively attack the competence of democracies, particularly the United States, its only superpower rival. China has also advanced internet governance concepts like “cyber sovereignty,” which is essentially an updated version of a long-standing authoritarian position that repression within sovereign, now cyber borders, should not be criticized by external actors based on international human rights law and principles.

Finally, we cannot ignore the fact that China’s growing global influence started with massive strategic investment in emerging technology. China recognized very early that dominance in technology would translate into power across other realms – military, economic, geopolitical, normative. The Chinese Communist Party publicly committed to win the AI-race against the US by 2030 and has made deep investments in other emerging technologies, such as quantum computing. Their current push to be first in developing a sovereign digital currency is another manifestation of their sophisticated understanding of the linkage between technology prominence and global power.

The bottom line: The Chinese government is on a mission to remake the 21st century global order in its own image and in accordance with its own repressive values. Democratic governments and stakeholders need to recognize the existential threat posed by this competing digital authoritarian model.

Solutions: How to Combat the Digital Authoritarian Model

Democratic stakeholders must move past trepidation about the scale and complexity of the threat and push toward practical action. As a first step, we need to recognize that we can’t beat something with nothing. While authoritarians have capitalized on AI and digital technologies for repression and control, democratic governments have been caught off-guard and without a compelling alternative vision of digital society that reflects democratic values. These are two sides of the same problem: without a proactive democratic plan, we are defenseless against an energetic digital authoritarian approach.

Admittedly, digitization and the AI-systems it has spawned have dramatically altered the context in which democracies must operate and human rights must be protected. But it is time for democratic leaders to face this reality and proactively engage in this radically changed digital environment with a human rights-based approach. The aggressive spread of authoritarian applications, concepts and practices must be met with a more attractive democratic alternative. In effect, a good offense will be our best defense.

The most promising prescriptions fall into three buckets of action.

1. Develop and Showcase a Democratic Model of Digital Restraint at Home

First, democracies need to get their own digital technology policies and practices in order. When it comes to government use of data and digital tools, rather than proactive values-based leadership, we have witnessed an unconscious drift toward unrestrained access to data and unchecked practices that violate privacy and civil liberties and undermine the rule of law. A significant part of the problem has been inadequate recognition, within government or the private sector, of the centrality of privacy to the exercise of all fundamental freedoms. In addition, the process principles of necessity, proportionality, and legality, which are essential features of the existing international human rights law framework and provide the basis for assessing the legitimacy of infringements on fundamental rights, have been ignored, in the main, in law enforcement and surveillance practices. Simply put, democratic governments have not developed adequate institutional constraints on their own use of data and AI-tools to sufficiently differentiate their practices from those of authoritarians.

Similarly, when it comes to regulatory efforts to combat “harmful” online content, many democracies see AI-scanning and filtering requirements as a silver bullet, without assessing the necessity and efficacy of those tools, or the proportionality of the infringement on liberty to the harms they seek to prevent. Flawed platform regulations promulgated by democracies have been copied directly in authoritarian leaning contexts, leading to a downward global spiral on core commitments to free expression and access to information. This trend demonstrates that the practices and regulatory approaches relied upon by democracies in their domestic contexts have significant unintended effects on global norms. Democratic governments must come to appreciate how important adherence to the rule of law and to the process principles of necessity and proportionality matters in a globalized digital environment.

To address this situation, democratic governments should develop and commit to human rights impact assessment processes — with necessity, proportionality and legality assessments built in — for their own procurement, use and regulation of data, platforms, and AI-enabled tools. Given the significant impact private sector technology companies have on the enjoyment of fundamental rights (especially privacy, freedom of expression, freedom of assembly and association, and the right to democratic participation), democratic governments also should mandate that private sector companies engage in human rights impact assessments in their development and deployment of products and services.

2. Exercise Leadership in International Tech Diplomacy and Standard Setting:

Second, democracies must reassert leadership in international arenas where cyber norms, tech standards and governance processes are established. These are the arenas in which system rivalry is playing out. While authoritarians have flooded the zone of multilateral cyber diplomacy and international technology policy, democracies have retreated from leadership due to their own lack of clarity about what adherence to human rights law and principles entails in the digital context.

An essential point here: We do not need new principles. Rather, we need to reinforce confidence in the continued relevance and applicability of existing international human rights law framework — which is globally recognized and has the status of international law. That said, we do need to undertake the hard work of articulating how to adhere to our existing international human rights framework in practice, in a radically changed digital context. A global multistakeholder process must be initiated to resolve disagreements among democratic actors about the legitimacy of different data practices, platform regulations, and applications of AI-enabled tools. Rebuilding the democratic alliance around a shared understanding of how to adhere to international human rights law in digital society should be treated as a strategic priority, on par with cybersecurity and more traditional dimensions of national security.

3. Invest in Tech Innovation as a Means of Advancing Democratic Values

Last, but not least, democracies need to recognize that normative leadership and technological leadership go together. If our goal is to spread democratic values rather than authoritarian norms, we must lead in technological innovation, particularly in AI and quantum computing. Dominance in those realms will translate into leverage and influence in normative realms and tech standard setting bodies. In addition, we need to become far more proactive in exporting democratic digital infrastructure as part of our trade and economic development aid programs, rather than ceding the opportunity to China to embed values into digital infrastructure in the developing world.

* * *

In sum, domestic practices, international norms, and technology innovation and standards are intertwined with digital governance systems. The strength of the digital authoritarian model stems from the fact that these elements are working in tandem. Democracies must recognize these interdependencies and demonstrate leadership in all three realms, simultaneously and in a coordinated fashion.

The tech practices we showcase in our domestic context, the norms for which we advocate in international tech fora, and the investments we make in emerging technologies and democratic information infrastructure will be mutually reinforcing. If this complex set of tasks is embraced and tackled with the sense of urgency and purpose it deserves, a prosperous, secure democratic future can be solidified. These are the essential elements from which we can build a democratic digital society.