News

Alphabet cuts ties with Australian AI firm that helped train Bard and Google Search

  •  

Alphabet has cut contractual ties with Appen, the artificial intelligence data firm that helped train Google’s chatbot Bard, Google Search results and other AI products.

After a “strategic review process,” Alphabet notified Appen over the weekend of the termination, which will go into effect March 19, according to a filing from Appen. The company said it had “no prior knowledge of Google’s decision to terminate the contract.”

Alphabet accounted for roughly one-third of Appen’s revenue, meaning the decision to end the relationship will impact “at least two thousand subcontracted Alphabet workers,” according to a statement Monday from the Alphabet Workers Union.

Appen, based in Australia, has helped train AI models for a star-studded list of tech behemoths. Five customers — MicrosoftAppleMetaGoogle and Amazon — have in the past accounted for 80% of Appen’s revenue. Appen has a platform of about 1 million freelance workers in more than 170 countries.

In 2023, revenue from work with Alphabet totaled $82.8 million of Appen’s $273 million in sales for the year, according to Monday’s filing.

Despite Appen’s enviable client list and its nearly 30-year history, the company has struggled in recent years with a loss of customers, a string of executive departures and plummeting financials — even as generative AI tools increased demand for training data. Revenue dropped 30% in 2023, after declining 13% a year earlier, which the company attributed in part to “challenging external operating and macro conditions.”

In August 2020, Appen’s shares peaked at AU$42.44 ($27.08) on the Australian Securities Exchange, sending its market cap to the equivalent of $4.3 billion. Now, the stock is trading at around AU$0.28, down more than 99% since its peak.

Former employees, who asked not to be named for fear of retaliation, told CNBC in September that the company’s current struggle to pivot to generative AI reflects years of weak quality controls and a disjointed organizational structure.

Appen’s past work for tech companies has been on projects like evaluating the relevance of search results, helping AI assistants understand requests in different accents, categorizing e-commerce images using AI, and building out map locations of electric vehicle charging stations, according to public information and interviews conducted by CNBC.

Appen has also touted its work on search relevance for Adobe and on translation services for Microsoft, as well as in providing training data for lidar companies, security applications and automotive manufacturers. 

But large language models of today operate differently. The underlying LLMs behind OpenAI’s ChatGPT and Google’s Bard are scouring the digital universe to provide sophisticated answers and advanced images in response to simple text queries. Companies are spending far more on processors from Nvidia and less on Appen.

Google and Appen have had conflicts in the past, namely a dispute about wages. In 2019, Google said its contractors would need to pay their workers $15 an hour. Appen didn’t meet that requirement, according to public letters written by some workers.

In January 2023, after months of organizing, raises went into effect for Appen freelancers working on the Bard chatbot and other Google products. The rates went up to between $14 and $14.50 per hour.

But labor issues persisted. In June, Appen faced charges from the U.S. National Labor Relations Board after allegedly firing six freelancers who spoke out publicly about frustrations with workplace conditions. The workers were later reinstated

Appen wrote in Monday’s filing that it will focus on managing costs, turning the business around and providing customers with quality AI data.

“Appen will immediately adjust its strategic priorities following the notification of the Google contract termination and provide further details in its FY23 full year results on 27 February 2024,” the company wrote.

WATCH: Google’s ‘showing some muscle’ with Bard AI upgrade


This article was originally published on CNBC