Meta removes terms that raised concerns over AI training

Jakarta (ANTARA) — Meta has reportedly removed an update to its terms that had raised concerns and suspicions about data collection for training the artificial intelligence (AI) models developed by the company.

TechCrunch reported on Saturday (17 October) that Meta had recently updated the terms of use for its commercial services.

Under the revised terms, the company could “use any content from a domain, web page or URL (business account) to provide, develop, improve, research, train and maintain the integrity” of Meta products and services.

Legal experts interviewed by Nieman Lab described the wording as an “extraordinary point”, saying it could authorise Meta to collect data from any publisher’s website with a Meta business account and use the content to train the company’s AI models.

A Meta spokesperson, however, said the terms update did not reflect a change in how the company conducts web searches.

The Verge subsequently reported that the wording which had prompted concern and suspicion had been removed.

The company also said the removal did not change how Meta uses web crawlers.