Tech giant Apple has responded to certain allegations regarding its AI models, saying it takes precautions at every stage \u2014 including design, model training, feature development, and quality evaluation \u2014 to identify how its AI tools may be misused or lead to potential harm.<\/p>\n\n\n\n
The company said in a technical paper that it will continuously and proactively \u201cimprove our AI tools with the help of user feedback\u201d.<\/p>\n\n\n\n
It last month revealed Apple Intelligence that will offer several generative AI features in iOS, macOS and iPadOS software over the next few months.<\/p>\n\n\n\n
\u201cThe pre-training data set consists of\u2026 data we have licensed from publishers, curated publicly available or open-sourced datasets and publicly available information crawled by our web crawler, Applebot,\u201d Apple wrote.<\/p>\n\n\n\n
Given our focus on protecting user privacy, we note that no private Apple user data is included in the data mixture, the company added.<\/p>\n\n\n\n
According to the technical paper, training data for the Apple Foundation Models (AFM) was sourced in a \u201cresponsible\u201d way.<\/p>\n\n\n\n
Apple Intelligence is designed with the company\u2019s core values at every step and built on a foundation of industry-lead privacy protection.<\/p>\n\n\n\n
\u201cAdditionally, we have created Responsible AI principles to guide how we develop AI tools, as well as the models that underpin them,\u201d said the iPhone maker.<\/p>\n\n\n\n
The company further said that no private Apple user data is included in the data mixture.<\/p>\n\n\n\n
\u201cAdditionally, extensive efforts have been made to exclude profanity, unsafe material, and personally identifiable information from publicly available data. Rigorous decontamination is also performed against many common evaluation benchmarks,\u201d Apple elaborated.<\/p>\n\n\n\n
To train its AI models, the company crawl publicly available information using its web crawler, Applebot and \u201crespect the rights of web publishers to opt out of Applebot using standard robots.txt directives\u201d.<\/p>\n\n\n\n
\u201cWe take steps to exclude pages containing profanity and apply filters to remove certain categories of personally identifiable information (PII). The remaining documents are then processed by a pipeline which performs quality filtering and plain-text extraction,\u201d Apple emphasised. AGENCIES<\/p>\n","protected":false},"excerpt":{"rendered":"
Tech giant Apple has responded to certain allegations regarding its AI models, saying it takes precautions at every stage \u2014 including design, model training, feature development, and quality evaluation \u2014 to identify how its AI tools may be misused or lead to potential harm. The company said in a technical paper that it will continuously […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"jetpack_post_was_ever_published":false,"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"footnotes":"","_jetpack_memberships_contains_paid_content":false,"jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","enabled":false}}},"categories":[87],"tags":[],"jetpack_publicize_connections":[],"yoast_head":"\n