Using AI to Investigate the Supply-chain Companies Aggravates the Global Supply-chain Turmoil
“Our members are uncertain of what’s acceptable proof to overcome that assumption of forced labor,” Eugene Laney, president of the American Association of Exporters and Importers, was quoted as saying.
Over 2000 batches of goods has been seized at the American border and this number is increasing recently since Biden administration’s Uyghur Forced Labor Prevention Act (UFLPA) came into effect, because it’s too hard to prove that the goods are irrelevant to forced labor. It is a daunting task to find out the relationships among suppliers of various levels because China is the world’s largest manufacturing center.
Under the UFLPA, enterprises are blacklisted by the U.S. when their goods are involved in manufacturing, in whole or in part, in China’s Xinjiang Uyghur Autonomous Region (XUAR) or are produced by certain Chinese entities related to XUAR. This Act requires the enterprises to provide “by clear and convincing evidence” to prove the products “were not produced using forced labor” within thirty days. We have to say that both the scope of crackdown and the proving period lead to high cost of compliance that global companies have to face in the short term.
Evan Smith, CEO of Altana AI,aSupply Chain Big Data Analytics Company, says that according to his company’s calculation, about one million out of the approximately ten million companies globally which purchase, sell or manufacture physical items will be affected by the law enforcement actions. China is the largest manufacturing center in the world, especially in the cotton, textiles, tomato and polysilicon industries which can hardly decouple completely with XUAR. The Act in force has already put global supply chains at risk of turmoil, but now the U.S. intends to further accelerate this process.
Sources say that the U.S. is using AI analytics technology to detect the direct and indirect transactions between global supplies and XUAR enterprises through tracking and identifying the global supply chain data. The analysis shows the XUAR-related supply chain involves 938,991 companies, 785,415 first-tier trading relationships and 6,871,643 second-tier trading relationships. This causes panic of global enterprises as well as AI experts.
Analysts point out that the true intention of the U.S. CHIPS Act and UFLPA is to suppress suppliers from specific countries and guarantee American global competitiveness in core areas through legislation. When the Act took effect, the global supply chain originally needed a gradual adjustment period, but the U.S. implemented AI with an attempt to accelerate this process, forcing the global enterprises to make instant supply chain shifts, therefore, companies are forced to seek out more expensive alternatives in a very short period, the cost will be paid by consumers ultimately. The already serious inflation in the U.S. and European countries will possibly be intensified and spread worldwide including 590 separate industries and 183 countries. It is hard to imagine how big of an impact the imbalance of the global supply chain has on the economic situation which has already been in cold winter.
AI involvement in law enforcement has always been a controversial issue on legal and moral level.
According to BBC, Facebook’s AI algorithm identified black males as gorillas, hence Facebook apologized “this is obviously an unacceptable error.” The same challenge exists for AI to identify companies with supply chain risk. The supply chain is so large and complex in recent globalized world so the authenticity of data and accuracy of algorithm can hardly meet the standards required by traditional due diligence techniques.
A deeper concern is that human supervision can hardly intervene in the automated AI system. Kate Crawford reveals the natural setback of AI-the accountability gap, i.e. the lack of human involvement in decisions that cause harm in AI Systems as State Actors. AI is similar to a black box, whose judgment logic, process is probably only known by R&D personnel. Which should bear the economic losses when AI wrongly prohibits the entry of a company’s products, the big data analytics company or the government? We can predict that if an enterprise is misjudged by AI,it can hardly appeal and has to swallow the bitter pill itself.
Meanwhile, the U.S. has introduced a series of trade sanctions bills in recent years, but the legislation supervision of data protection and AI usage is still lagging behind. The big data analytics providers may make use of the U.S. dominance in trade to directly get access to the supply chain information such as specific addresses, cargo numbers, delivery and flow of goods and other information of a company worldwide, for economic strikes or other political purposes. When a country is a rule maker and judge, forcing global implementation of its standards, people have to fear what will be the next industry after the cotton and chips. It is certain that not only the sanctioned countries will bear the consequences, but also the enterprises and people worldwide, including the U.S., will pay the price.
The world has already faced the risk of economic collapse, the food and energy crises are emerging one after another under the influence of epidemics, wars and inflation. Alan Bersin, a former commissioner of U.S. Customs and Border Protection and the current executive chairman of Altana AI, said “the impact of this on the global economy and on the U.S. economy is measured in the many billions of dollars, not in the millions of dollars.” in an interview with the New York Times.
“The public is unprepared for what is about to happen”, he stressed.