UN Comes Out Against Some Artificial Intelligence
The United Nations (UN) urged a moratorium on Artificial Intelligence that could be harmful to human rights. Specifically, the UN High Commissioner for Human Rights, Michelle Bachelet criticized facial recognition and social scoring software. Both facial recognition software and social scoring software are closely associated with China. This week Curtin University in Perth, Australia demanded a retraction of one of its own studies funded by the Chinese government because the technology was subsequently used to identify Uighurs in China.
Fears of Artificial Intelligence misuse are rampant and sometimes well-founded. One need not believe that artificial intelligence is going to lead to Skynet or the Master Control Program to be worried that AI might exert undue influence in policy or the real world. In a recent especially unfortunate faux pas, a Facebook facial recognition program insultingly labeled Black people’s portraits as “primates.” Even more pernicious, AI might mask racial disparities in bank loans. Given the demonstrated risks of accidental disparate treatment, deliberately creating developing AI to repress groups could be a major threat to international human rights.
It is not obvious that the uses the UN condemns are AI, and in some instances the software the UN is concerned about may be Machine Learning (ML) or simple algorithms. I will not go in-depth about the differences between various computation tools that governments or companies might misuse to oppress people. Simple algorithms, rules that govern outcomes can be just as oppressive as AI or ML. AI and ML can inspire more fear because of how AI and ML generate their outcomes can be opaque, even to relatively well-informed observers.
Zoho Under Attack by Government
The Cybersecurity and Infrastructure Security Agency (CISA) issued a warning that the Zoho identity authentication tool has been under active attack by an Advanced Persistent Threat (APT) since August. Zoho is primarily a Customer Relationship Management suite, to help companies track contacts with customers, and respond to customer requests. As part of its suite, Zoho offers single sign-on (SSO) identity verification allowing other companies to use Zoho logins. Using Zoho SSO is like using your Google or Facebook account to log in to accounts other than Google or Facebook.
APTs are almost always state-sponsored organizations, and a government attacking a company so directly is alarming. Outsourcing user authentication is a common practice for companies and is usually the more secure way to manage online identities. A government attacking identity management like this compromises multiple companies, many of which lack the resources to meet an APT threat. Zoho has already patched the vulnerability, but the CISA be a warning to smaller organizations that they are in some government’s cross-hairs.
AUKUS Is a Game Changer
The biggest story in international relations this past week was the announcement of the AUKUS pact, and I will address the pact in a separate essay later this week. The AUKUS pact is likely to go down as an early turning point in this decade’s international relations; it is a big freaking deal. The primary effect of the agreement is to join the US, UK, and Australia in a security pact, which is probably only second to NATO in terms of military capability. The secondary effect is that the US will now build nuclear-powered submarines for Australia.
There are winners and losers in this deal, and the effects are still being felt. The People’s Republic of China is livid―and Taiwan is ecstatic―because the PRC views the alliance as a check on Chinese ambitions in the Asia Pacific. France is fuming because Australia is reneging on a $50 billion submarine contract, but I suspect the relationship will survive. New Zealand is in a tight spot, because New Zealand has a long history of opposition to nuclear power, but is also under nearly relentless cyberattacks, and is likely to come under additional pressure to align with one side or the other.
David Benson is a Professor of Strategy and National Security focusing on cyberstrategy and international relations. You can reach him at dbenson@osiriscodex.com.
To get more insightful analysis like this in your inbox at no cost please subscribe.
Ask a question! Raise an objection! Leave a comment!