Lawmakers Investigate Use of Artificial Intelligence ‘Ghost’ Employees by Amazon, Alphabet, Microsoft, and Meta

Lawmakers from the Democratic party are urging leading technology companies to provide transparency regarding the working conditions of their “ghost workers.” These invisible laborers, such as individuals responsible for data labeling and rating responses, have become crucial to the rapid growth of artificial intelligence.

“Despite the essential nature of this work, millions of data workers around the world perform these stressful tasks under constant surveillance, low pay and no benefits,” a group of lawmakers led by Massachusetts Sen. Ed Markey and Washington Rep. Pramila Jayapal wrote in a letter Wednesday to the CEOs of nine companies, including Amazon.com Inc., Alphabet Inc., Meta Platforms Inc., Microsoft Corp. and International Business Machines Corp.

“Employees are expected to screen out unsafe chatbot responses, but may have little time to assess the safety of the response,” they added. “Data workers are often given little training or oversight, which can lead to bias.”

Lawmakers are asking executives to answer broad questions about their data workforce, including their ability to take breaks, appeal suspensions or access mental health resources when faced with traumatizing content. “Technology companies must not build AI on the backs of exploited workers,” wrote Democrats, including Massachusetts senator and former presidential candidate Elizabeth Warren.

In addition to established technology giants, the recipients of the letter include newer companies focused on artificial intelligence OpenAI Inc., Inflection AI, Scale AI Inc. and Anthropic.

U.S. companies rely heavily on subcontractors, located domestically or abroad, who are hired through third-party staffing firms and often do not have the benefits offered to the companies’ own direct employees, to develop AI products. Companies also use this kind of work in other heavy tasks, such as content moderation and product quality assurance.

Generative AI tools, which produce responses to text prompts in the form of text, photos or even videos, typically hire thousands of unprecedented contract workers to train, fix and improve algorithms that are then presented to customers as if they worked by magic. about technology. They are often underpaid and say they are stressed and overworked.

In some cases, workers report trauma from viewing disturbing images to filter them out. OpenAI paid Kenyan workers less than $2 an hour to keep such content off ChatGPT, Time reported in January. Google’s ChatGPT competitor Bard also relies on human labor, which told Bloomberg they are given little training, high workloads and poor pay.

The lawmakers’ letter comes as executives, including those from Tesla Inc., Meta, Microsoft and Alphabet, meet with senators Wednesday afternoon at a closed-door AI summit hosted by Senate Majority Leader Chuck Schumer, who was not among the document’s signatories. .

“These tech moguls are underpaying workers, denying them basic protections and benefits, and subjecting them to a vast web of surveillance to support their business,” Markey said via email. “When they come to the Capitol to tout their innovation and excellence, I’d like to hear them answer for these abhorrent labor practices.”