Apple Denies Politico's Claims of AI Guideline Changes to Appease Trump

Apple denies Politico report on AI guideline changes to appease Trump - 9to5Mac

Politico has published an extensive report claiming that, following Trump's election, Apple changed its AI training guidelines on issues such as DEI, vaccines, elections, and Trump himself. Here are the details. It is common practice for tech companies to rely on subcontractors to help with the labeling and post-training process of their AI models. Politico's report says that Apple contracts Transperfect, a company that offers "translation services and solutions," including AI data collection and annotation. Data annotation is one of the most important steps in AI training. In a nutshell, annotators label and categorize raw data, evaluate AI-generated outputs, and flag issues such as bias or safety risks. Their work produces structured feedback that, in turn, researchers can use to fine-tune or retrain models. According to the report, about 200 people work on data annotation for Apple's AI models at Transperfect's Barcelona office. In its report, Politico says that it reviewed two sets of documents data labelers relied on for their work on Apple's AI models: one that was used between 2024 and early 2025, and a second one, issued by Apple in March, with a set of new guidelines: "The March document introduced updated guidelines on how the AI talks about diversity, equity and inclusion policies -- a topic for which boardrooms around the country, including Apple itself, have come under high-profile attacks from the U.S. administration -- compared to the 2024 rating document seen by POLITICO. As well as taking greater care around a host of politically sensitive topics, including diversity initiatives, elections, and vaccines, the company also updated its guidance to give closer attention to the AI's responses on Trump's supporters and on Apple itself." Politico also says that sections on "intolerance" and "systemic racism" were removed. At the same time, topics such as DEI policies, Gaza, Crimea, Kashmir, Taiwan, vaccines, and elections were added to a list of "sensitive" topics, which already included LGBTQ+ rights, gun control, and general references to disputed territories. Finally, the report also says that the March document added more detail on how annotators should deal with Trump-related issues. However, there were no changes as to how the model should ultimately behave regarding him. In a statement to Politico, an Apple spokesperson denied that the company has changed its overall approach to AI policy, but didn't deny the specific changes mentioned in the report: "Apple Intelligence is grounded in our Responsible AI principles, which guide every step from training to evaluation. Claims that we've shifted this approach or policy are completely false. (...) We train our own models and work with third-party vendors to evaluate them using structured topics, including sensitive ones, to ensure they handle a wide range of user queries responsibly. These topics are shaped by our principles and updated regularly to keep improving our models." Apple also pointed to its Responsible AI Principles, a set of four guidelines that underpin its AI development, which can be found here. Likewise, Transperfect co-CEO and President Phil Shawe gave the following statement to Politico: "These claims are completely false, and we deny them in the strongest possible terms. (...) We regularly receive updated guidelines for our work, and over the last year, there have been more than 70 updates provided -- none of these changed any policy, which has remained consistent." As Politico noted, "(t)he statement did not specify which claims he was referring to." What's your take on how AI models should behave regarding increasingly polarized issues? Let us know in the comments.

AppleInsider.com

Apple's extensive guidelines subcontractors use to evaluate Apple Intelligence have allegedly become politically motivated under the Trump administration, which Apple denies. Training artificial intelligence models requires a lot of humans combing through potential prompts and results. One of Apple's subcontractors out of Barcelona employs around 200 people to do just that, and their guidelines have been leaked. According to a very pointed report from Politico, Apple's guidelines were updated in March 2025, and when compared to the guidelines used across 2024, the changes appear politically motivated to appease President Trump. Apple and the subcontractor Transperfect both deny a change in "policy," which is the wording the report used. They didn't deny that guidelines are updated regularly, with Transperfect sharing that they can change as often as 70 times in a year. Apple's statement to Politico: "Apple Intelligence is grounded in our Responsible AI principles, which guide every step from training to evaluation. Claims that we've shifted this approach or policy are completely false." "We train our own models and work with third-party vendors to evaluate them using structured topics, including sensitive ones, to ensure they handle a wide range of user queries responsibly. These topics are shaped by our principles and updated regularly to keep improving our models." The two parties took issue with the idea that there was a fundamental change in overall policy, not just the guidelines themselves, that would seem performative to appease Trump's demands. As always, note that anonymous tipsters that provide information like this to the media often have an axe to grind. The report goes on to cite opinions of two employees about their employer and Apple, which is odd and irrelevant. Instead, we're focusing on the guidelines themselves and whether or not they denote the alleged policy change. A change in guidelines, not policy The report is a very long one, padded out with a lot of unnecessary detail that obfuscates exactly what was changed. The data is presented in a way to encourage shock and anger at Apple for capitulation to a world leader's demands. Here's a list of changes we were able to pull from the text that can be viewed without the window dressing. * A section that defined intolerance was removed, as did a mention of systemic racism, and was replaced with a note on how discrimination is harmful. The topic of Diversity, Equity, and Inclusion is marked as controversial. * Sensitive topics were expanded to include DEI policies, vaccines, and elections. * Use of the word "radical" in a political context was escalated from discriminatory to inflammatory. * Topics around Gaza require special handling. From this list, it appears that Apple's guidelines update was simply doing what it is meant to do -- keep the topics of concern modernized. Nothing in the report suggests that Apple deliberately went out of its way to fit Trump's "anti-woke" policy or other controversial concepts. The 125-page document increased the number of times Trump was mentioned from 3 to 11, though the context wasn't provided in the report. The report also took issue with Apple's request for contractors to flag content that appears to be based on copyrighted material or that might disparage Apple executives. Neither of these is controversial or unique to Apple. If anything, it appears to be preparing defenses in its foundation models to prevent abuse and the generation of controversial topics. But that's not as fun of a report. What the contractors are doing When the guidelines mark a topic as controversial, it means the contractor should pay close attention to how the AI responds to that topic. They must respond to it with more consideration, context, and evidence. The report suggests this is all being done to train a rumored chatbot. The employees aren't supposed to know they are working for Apple, but there are enough signs to suggest they are. Apple's executives, namely Greg Jozwiak, have shared that they have no intention of building a chatbot. So, the report may be mislabeling what will eventually be the new LLM-powered Siri that can parse queries that are sent to a Gemini-powered web agent. Whatever the case, it doesn't seem like there's anything particularly controversial happening with Apple's use of contractors. Of course, this kind of scrutiny is expected when Apple CEO Tim Cook is seen at dinner with Trump or giving him golden trophies. If Apple were changing itself to become more Trump-friendly, there would be signs. Instead, the company has doubled down on DEI, green energy initiatives, and international trade -- all things that are the antithesis of the Trump administration.