Google terminated its contract with Appen, an Australian data company that played a role in training its extensive language model AI tools for products like Bard and Search. This decision comes amidst the growing competition in the development of generative AI tools. Courtenay Mencini, a spokesperson for Google, stated that the contract termination was part of an ongoing effort to assess and optimize supplier partnerships within Alphabet for enhanced operational efficiency.
Appen, in a filing with the Australian Securities Exchange, expressed surprise at Google’s abrupt contract termination, indicating no prior knowledge of the decision. Human workers at companies such as Appen often undertake the less desirable aspects of AI training, serving as the essential yet often overlooked workforce within the industry. These workers assist in evaluating data quality and AI model responses.
Reports indicate that Appen has also contributed to training AI models for Microsoft, Meta, and Amazon. The company acknowledged the significant impact of its collaboration with Google on its revenue, with earnings from Google alone reaching $82.8 million in the fiscal year 2023. Despite revenue growth, Appen faced challenges, leading to layoffs due to business conditions.
In a related development, employees at Accenture, another Google contractor, voted overwhelmingly to join the Alphabet Workers Union after refusing to engage with objectionable content for the Bard chatbot. This move underscores the increasing concerns surrounding the nature of work handled by AI contractors.
Beyond Google, issues of fair compensation and ethical practices persist in the tech industry. Content moderators in Kenya, employed by data-labeling firm Sama, filed a lawsuit against the company and its client Meta for allegedly paying workers a meager $2.20 per hour to review disturbing images and videos, highlighting broader challenges within the sector.
