Do you think engaging with an emerging tech tool can change your firmly held beliefs? Or sway you toward a decision you would not have otherwise made? Most of us humans think we are too smart for that, but mounting evidence suggests otherwise.
When it comes to a new crop of generative artificial intelligence (AI) technology, the power of “persuasion” has been identified as a potentially catastrophic risk right alongside fears that models could gain autonomy or help build a nuclear weapon. Separately, lower-stakes designs meant to influence behavior are already ubiquitous in the products many of us use everyday, nudging us to endlessly scroll on social platforms, or open Snapchat or Duolingo to continue a “streak.”
However, recent advances in that nascent technology from China are raising fresh national security concerns. New research funded by the US Department of State and released by an Australian think tank found that Chinese tech companies are on the cusp of creating and deploying technologies with “unprecedented persuasive capabilities.”
Illustration: Yusha
From a security perspective, that could be abused by Beijing or other actors to sway political opinions or sow social unrest and division. In other words, it is a weapon to subdue enemies without any fighting, the war tactic heralded by the Chinese philosopher General Sun Zi (孫子).
The Australian Strategic Policy Institute report published last week identified China’s commercial sector as “already a global leader” in the development and adoption of products designed to change attitudes or behaviors by exploiting physiological or cognitive vulnerabilities. To accomplish that, the tools rely heavily on analyzing personal data they collect and then tailor interactions with users. The paper identified a handful of Chinese firms that it says are already using such technology — spanning generative AI, virtual reality and the more emerging neurotechnology sector — to support Beijing’s propaganda and military goals.
However, that is also very much a global issue. China’s private sector might be racing ahead to develop persuasive methods, but it is following playbooks developed by US’ big tech firms to better understand their users and keep them engaged. Addressing the Beijing risk would require us to properly unpack how we let tech products influence our lives. However, fresh national security risks, combined with how AI and other new innovations can quickly scale up these tools’ effectiveness, should be a wake-up call at a time when persuasion is already so entrenched into Silicon Valley product design.
Part of what makes addressing this issue so difficult is that it can be a double-edged sword. A science study published earlier this year found that chatting with AI models could convince conspiracy theorists to reduce their beliefs, even among those who said they were important to their identity. That highlighted the positive “persuasive powers” of large language models and their ability to engage with personalized dialogue, the researchers said.
How to prevent those powers from being employed by Beijing or other bad actors for nefarious campaigns would be an increasing challenge for policymakers that goes beyond cutting off access to advanced semiconductors.
Demanding far more transparency would be one way to start, by requiring tech companies to provide clear disclosures when content is tailored in a way that could influence behaviors. Expanding data protection laws or giving users clearer ways to opt-out of having their information collected would also limit the ability of those tools to individually target users.
Prioritizing digital literacy and education is also imperative to raise awareness about persuasive technologies, how algorithms and personalized content work, how to recognize tactics and how to avoid being potentially manipulated by these systems.
Ultimately, a lot more research is needed on how to protect people from the risks of persuasive technology and it would be wise for the companies behind these tools to lead the charge, as firms such as OpenAI and Anthropic have begun doing with AI. Policymakers should also demand firms share findings with regulators and relevant stakeholders to build a global understanding of how those techniques could be exploited by adversaries. That information could then be used to set clear standards or targeted regulation.
The risk of technology so sophisticated that it allowing Beijing to pull the strings to change what you believe or who you are might still seem like a far-off, sci-fi concern. However, the stakes are too high for global policymakers to respond only after that has been unleashed. Now is the time for a global reckoning on how much personal information and influence we give tech companies over our lives.
Catherine Thorbecke is a Bloomberg Opinion columnist covering Asia tech. Previously she was a tech reporter at CNN and ABC News.
It is almost three years since Chinese President Xi Jinping (習近平) and Russian President Vladimir Putin declared a friendship with “no limits” — weeks before Russia’s invasion of Ukraine in February 2022. Since then, they have retreated from such rhetorical enthusiasm. The “no limits” language was quickly dumped, probably at Beijing’s behest. When Putin visited China in May last year, he said that he and his counterpart were “as close as brothers.” Xi more coolly called the Russian president “a good friend and a good neighbor.” China has conspicuously not reciprocated Putin’s description of it as an ally. Yet the partnership
The ancient Chinese military strategist Sun Tzu (孫子) said “know yourself and know your enemy and you will win a hundred battles.” Applied in our times, Taiwanese should know themselves and know the Chinese Communist Party (CCP) so that Taiwan will win a hundred battles and hopefully, deter the CCP. Taiwanese receive information daily about the People’s Liberation Army’s (PLA) threat from the Ministry of National Defense and news sources. One area that needs better understanding is which forces would the People’s Republic of China (PRC) use to impose martial law and what would be the consequences for living under PRC
Chinese Nationalist Party (KMT) Chairman Eric Chu (朱立倫) said that he expects this year to be a year of “peace.” However, this is ironic given the actions of some KMT legislators and politicians. To push forward several amendments, they went against the principles of legislation such as substantive deliberation, and even tried to remove obstacles with violence during the third readings of the bills. Chu says that the KMT represents the public interest, accusing President William Lai (賴清德) and the Democratic Progressive Party of fighting against the opposition. After pushing through the amendments, the KMT caucus demanded that Legislative Speaker
Although former US secretary of state Mike Pompeo — known for being the most pro-Taiwan official to hold the post — is not in the second administration of US president-elect Donald Trump, he has maintained close ties with the former president and involved himself in think tank activities, giving him firsthand knowledge of the US’ national strategy. On Monday, Pompeo visited Taiwan for the fourth time, attending a Formosa Republican Association’s forum titled “Towards Permanent World Peace: The Shared Mission of the US and Taiwan.” At the event, he reaffirmed his belief in Taiwan’s democracy, liberty, human rights and independence, highlighting a
RSS