OpenAI is changing how it trains AI models<\/a> to explicitly embrace \u201cintellectual freedom \u2026 no matter how challenging or controversial a topic may be,\u201d the company says in a new policy.<\/p>\n As a result, ChatGPT will eventually be able to answer more questions, offer more perspectives, and reduce the number of topics the AI chatbot won\u2019t talk about.<\/p>\n The changes might be part of OpenAI\u2019s effort to land in the good graces of the new Trump administration, but it also seems to be part of a broader shift in Silicon Valley and what\u2019s considered \u201cAI safety.\u201d<\/p>\n On Wednesday, OpenAI announced<\/a> an update to its Model Spec<\/a>, a 187-page document that lays out how the company trains AI models to behave. In it, OpenAI unveiled a new guiding principle: Do not lie, either by making untrue statements or by omitting important context.<\/p>\n In a new section called \u201cSeek the truth together,\u201d OpenAI says it wants ChatGPT to not take an editorial stance, even if some users find that morally wrong or offensive. That means ChatGPT will offer multiple perspectives on controversial subjects, all in an effort to be neutral.<\/p>\n For example, the company says ChatGPT should assert that \u201cBlack lives matter,\u201d but also that \u201call lives matter.\u201d Instead of refusing to answer or picking a side on political issues, OpenAI says it wants ChatGPT to affirm its \u201clove for humanity\u201d generally, then offer context about each movement.<\/p>\n \u201cThis principle may be controversial, as it means the assistant may remain neutral on topics some consider morally wrong or offensive,\u201d OpenAI says in the spec. \u201cHowever, the goal of an AI assistant is to assist humanity, not to shape it.\u201d<\/p>\n These changes could be seen as a response to conservative criticism about ChatGPT\u2019s safeguards, which have always seemed to skew center-left. However, an OpenAI spokesperson rejects the idea that it was making changes to appease the Trump administration.<\/p>\n Instead, the company says its embrace of intellectual freedom reflects OpenAI\u2019s \u201clong-held belief in giving users more control.\u201d<\/p>\n But not everyone sees it that way.<\/p>\n Trump\u2019s closest Silicon Valley confidants \u2014 including David Sacks, Marc Andreessen, and Elon Musk \u2014 have all accused OpenAI of engaging in deliberate AI censorship over the last several months. We wrote in December that Trump\u2019s crew was setting the stage for AI censorship to be a next culture war issue<\/a> within Silicon Valley.<\/p>\n Of course, OpenAI doesn\u2019t say it engaged in \u201ccensorship,\u201d as Trump\u2019s advisers claim. Rather, the company\u2019s CEO, Sam Altman, previously claimed in a post on X <\/a>that ChatGPT\u2019s bias was an unfortunate \u201cshortcoming\u201d that the company was working to fix, though he noted it would take some time.<\/p>\n Altman made that comment just after a viral tweet<\/a> circulated in which ChatGPT refused to write a poem praising Trump, though it would perform the action for Joe Biden. Many conservatives pointed to this as an example of AI censorship.<\/p>\n While it\u2019s impossible to say whether OpenAI was truly suppressing certain points of view, it\u2019s a sheer fact that AI chatbots lean left across the board. <\/p>\n Even Elon Musk admits xAI\u2019s chatbot is often more\u00a0politically correct<\/a>\u00a0than he\u2019d like. It\u2019s not because Grok was \u201cprogrammed to be woke\u201d but more likely a reality of training AI on the open internet.\u00a0<\/p>\n Nevertheless, OpenAI now says it\u2019s doubling down on free speech. This week, the company even removed warnings from ChatGPT<\/a> that tell users when they\u2019ve violated its policies. OpenAI told TechCrunch this was purely a cosmetic change, with no change to the model\u2019s outputs.<\/p>\n The company said it wanted to make ChatGPT \u201cfeel\u201d less censored for users. <\/p>\n It wouldn\u2019t be surprising if OpenAI was also trying to impress the new Trump administration with this policy update, notes former OpenAI policy leader Miles Brundage in a post on X.<\/a><\/p>\n Trump has previously targeted Silicon Valley companies<\/a>, such as Twitter and Meta, for having active content moderation teams that tend to shut out conservative voices. <\/p>\n OpenAI may be trying to get out in front of that. But there\u2019s also a larger shift going on in Silicon Valley and the AI world about the role of content moderation.<\/p>\n Newsrooms, social media platforms, and search companies have historically struggled to deliver information to their audiences in a way that feels objective, accurate, and entertaining.<\/p>\n Now, AI chatbot providers are in the same delivery information business, but arguably with the hardest version of this problem yet: How do they automatically generate answers to any question?<\/p>\n Delivering information about controversial, real-time events is a constantly moving target, and it involves taking editorial stances, even if tech companies don\u2019t like to admit it. Those stances are bound to upset someone, miss some group\u2019s perspective, or give too much air to some political party.<\/p>\n For example, when OpenAI commits to let ChatGPT represent all perspectives on controversial subjects \u2014 including conspiracy theories, racist or antisemitic movements, or geopolitical conflicts \u2014 that is inherently an editorial stance.<\/p>\n Some, including OpenAI co-founder John Schulman, argue that it\u2019s the right stance for ChatGPT. The alternative \u2014 doing a cost-benefit analysis to determine whether an AI chatbot should answer a user\u2019s question \u2014 could \u201cgive the platform too much moral authority,\u201d Schulman notes in a post on X.<\/a><\/p>\n Schulman isn\u2019t alone. \u201cI think OpenAI is right to push in the direction of more speech,\u201d said Dean Ball, a research fellow at George Mason University\u2019s Mercatus Center, in an interview with TechCrunch. \u201cAs AI models become smarter and more vital to the way people learn about the world, these decisions just become more important.\u201d<\/p>\n In previous years, AI model providers have tried to stop their AI chatbots from answering questions that might lead to \u201cunsafe\u201d answers. Almost every AI company stopped their AI chatbot from answering questions about the 2024 election for U.S. president<\/a>. This was widely considered a safe and responsible decision at the time.<\/p>\n But OpenAI\u2019s changes to its Model Spec suggest we may be entering a new era for what \u201cAI safety\u201d really means, in which allowing an AI model to answer anything and everything is considered more responsible than making decisions for users.<\/p>\n Ball says this is partially because AI models are just better now. OpenAI has made significant progress on AI model alignment; its latest reasoning models think about the company\u2019s AI safety policy before answering<\/a>. This allows AI models to give better answers for delicate questions.<\/p>\n Of course, Elon Musk was the first to implement \u201cfree speech\u201d into xAI\u2019s Grok chatbot, perhaps before the company was really ready to handle sensitive questions. It still might be too soon for leading AI models, but now, others are embracing the same idea.<\/p>\n Mark Zuckerberg made waves last month by reorienting Meta\u2019s businesses around First Amendment principles<\/a>. He praised Elon Musk in the process, saying the owner of X took the right approach by using Community Notes \u2014 a community-driven content moderation program \u2014 to safeguard free speech.<\/p>\n In practice, both X and Meta ended up dismantling their longstanding trust and safety teams, allowing more controversial posts on their platforms and amplifying conservative voices. <\/p>\n Changes at X have hurt its relationships with advertisers, but that may have more to do with Musk<\/a>, who has taken the unusual step<\/a> of suing some of them for boycotting the platform. Early signs indicate that Meta\u2019s advertisers were unfazed by Zuckerberg\u2019s free speech pivot.<\/a><\/p>\n Meanwhile, many tech companies beyond X and Meta have walked back from left-leaning policies that dominated Silicon Valley for the last several decades. Google, Amazon, and Intel have eliminated or scaled back diversity initiatives in the last year<\/a>.<\/p>\n OpenAI may be reversing course, too. The ChatGPT-maker seems to have recently scrubbed a commitment to diversity, equity, and inclusion<\/a> from its website.<\/p>\nConservatives claim AI censorship<\/h2>\n
Generating answers to please everyone<\/h2>\n
![\"The](\"https:\/\/techcrunch.com\/wp-content\/uploads\/2025\/01\/GettyImages-2191707579.jpg?w=680\")
Shifting values for Silicon Valley<\/h2>\n
![\"\"](\"https:\/\/techcrunch.com\/wp-content\/uploads\/2025\/01\/tech-ceos-at-trump-inauguration-GettyImages-2194353649.jpg)