OpenAI Quietly Deletes Ban on Using ChatGPT for Military and Warfare – The Intercept
OpenAI this week quietly deleted language expressly prohibiting the use of its technology for military purposes from its usage policy, which seeks to dictate how powerful and immensely popular tools like ChatGPT can be used.
Up until January 10, OpenAIs usage policies page included a ban on activity that has high risk of physical harm, including, specifically, weapons development and military and warfare. That plainly worded prohibition against military applications would seemingly rule out any official, and extremely lucrative, use by the Department of Defense or any other state military. The new policy retains an injunction not to use our service to harm yourself or others and gives develop or use weapons as an example, but the blanket ban on military and warfare use has vanished.
The unannounced redaction is part of a major rewrite of the policy page, which the company said was intended to make the document clearer and more readable, and which includes many other substantial language and formatting changes.
We aimed to create a set of universal principles that are both easy to remember and apply, especially as our tools are now globally used by everyday users who can now also build GPTs, OpenAI spokesperson Niko Felix said in an email to The Intercept. A principle like Dont harm others is broad yet easily grasped and relevant in numerous contexts. Additionally, we specifically cited weapons and injury to others as clear examples.
Felix declined to say whether the vaguer harm ban encompassed all military use, writing, Any use of our technology, including by the military, to [develop] or [use] weapons, [injure] others or [destroy] property, or [engage] in unauthorized activities that violate the security of any service or system, is disallowed.
OpenAI is well aware of the risk and harms that may arise due to the use of their technology and services in military applications, said Heidy Khlaaf, engineering director at the cybersecurity firm Trail of Bits and an expert on machine learning and autonomous systems safety, citing a 2022 paper she co-authored with OpenAI researchers that specifically flagged the risk of military use. Khlaaf added that the new policy seems to emphasize legality over safety. There is a distinct difference between the two policies, as the former clearly outlines that weapons development, and military and warfare is disallowed, while the latter emphasizes flexibility and compliance with the law, she said. Developing weapons, and carrying out activities related to military and warfare is lawful to various extents. The potential implications for AI safety are significant. Given the well-known instances of bias and hallucination present within Large Language Models (LLMs), and their overall lack of accuracy, their use within military warfare can only lead to imprecise and biased operations that are likely to exacerbate harm and civilian casualties.
The real-world consequences of the policy are unclear. Last year, The Intercept reported that OpenAI was unwilling to say whether it would enforce its own clear military and warfare ban in the face of increasing interest from the Pentagon and U.S. intelligence community.
Given the use of AI systems in the targeting of civilians in Gaza, its a notable moment to make the decision to remove the words military and warfare from OpenAIs permissible use policy, said Sarah Myers West, managing director of the AI Now Institute and a former AI policy analyst at the Federal Trade Commission. The language that is in the policy remains vague and raises questions about how OpenAI intends to approach enforcement.
While nothing OpenAI offers today could plausibly be used to directly kill someone, militarily or otherwise ChatGPT cant maneuver a drone or fire a missile any military is in the business of killing, or at least maintaining the capacity to kill. There are any number of killing-adjacent tasks that a LLM like ChatGPT could augment, like writing code or processing procurement orders. A review of custom ChatGPT-powered bots offered by OpenAI suggests U.S. military personnel are already using the technology to expedite paperwork. The National Geospatial-Intelligence Agency, which directly aids U.S. combat efforts, has openly speculated about using ChatGPT to aid its human analysts. Even if OpenAI tools were deployed by portions of a military force for purposes that arent directly violent, they would still be aiding an institution whose main purpose is lethality.
Experts who reviewed the policy changes at The Intercepts request said OpenAI appears to be silently weakening its stance against doing business with militaries. I could imagine that the shift away from military and warfare to weapons leaves open a space for OpenAI to support operational infrastructures as long as the application doesnt directly involve weapons development narrowly defined, said Lucy Suchman, professor emerita of anthropology of science and technology at Lancaster University. Of course, I think the idea that you can contribute to warfighting platforms while claiming not to be involved in the development or use of weapons would be disingenuous, removing the weapon from the sociotechnical system including command and control infrastructures of which its part. Suchman, a scholar of artificial intelligence since the 1970s and member of the International Committee for Robot Arms Control, added, It seems plausible that the new policy document evades the question of military contracting and warfighting operations by focusing specifically on weapons.
Suchman and Myers West both pointed to OpenAIs close partnership with Microsoft, a major defense contractor, which has invested $13 billion in the LLM maker to date and resells the companys software tools.
The changes come as militaries around the world are eager to incorporate machine learning techniques to gain an advantage; the Pentagon is still tentatively exploring how it might use ChatGPT or other large-language models, a type of software tool that can rapidly and dextrously generate sophisticated text outputs. LLMs are trained on giant volumes of books, articles, and other web data in order to approximate human responses to user prompts. Though the outputs of an LLM like ChatGPT are often extremely convincing, they are optimized for coherence rather than a firm grasp on reality and often suffer from so-called hallucinations that make accuracy and factuality a problem. Still, the ability of LLMs to quickly ingest text and rapidly output analysis or at least the simulacrum of analysis makes them a natural fit for the data-laden Defense Department.
While some within U.S. military leadership have expressed concern about the tendency of LLMs to insert glaring factual errors or other distortions, as well as security risks that might come with using ChatGPT to analyze classified or otherwise sensitive data, the Pentagon remains generally eager to adopt artificial intelligence tools. In a November address, Deputy Secretary of Defense Kathleen Hicks stated that AI is a key part of the comprehensive, warfighter-centric approach to innovation that Secretary [Lloyd] Austin and I have been driving from Day 1, though she cautioned that most current offerings arent yet technically mature enough to comply with our ethical AI principles.
Last year, Kimberly Sablon, the Pentagons principal director for trusted AI and autonomy, told a conference in Hawaii that [t]heres a lot of good there in terms of how we can utilize large-language models like [ChatGPT] to disrupt critical functions across the department.
Read more:
OpenAI Quietly Deletes Ban on Using ChatGPT for Military and Warfare - The Intercept
- The people refusing to use AI - BBC - May 5th, 2025 [May 5th, 2025]
- Trump posts AI image of himself as pope, leaving Catholics offended and unamused as conclave nears - CNN - May 5th, 2025 [May 5th, 2025]
- The Deadly AI Slow Roll in SaaS: It May Cost You Everything - SaaStr - May 5th, 2025 [May 5th, 2025]
- President Trump shares AI-generated photo of himself dressed as pope - CBS News - May 5th, 2025 [May 5th, 2025]
- 'They can't take a joke': Trump says he knew 'nothing' about AI image of him as the pope - USA Today - May 5th, 2025 [May 5th, 2025]
- 2 Magnificent Artificial Intelligence (AI) Stocks to Buy in May and 1 to Avoid - The Motley Fool - May 5th, 2025 [May 5th, 2025]
- Catholic community reacts to Trump's AI image of himself as the pope - ABC News - May 5th, 2025 [May 5th, 2025]
- Palantir raises annual revenue forecast on AI demand but investors unimpressed - Reuters - May 5th, 2025 [May 5th, 2025]
- The new IT stack: Rebuilding infrastructure for an AI-first world - cio.com - May 5th, 2025 [May 5th, 2025]
- Function Health acquires Ezra to combine lab testing and AI-powered medical imaging for preventive health - Fierce Healthcare - May 5th, 2025 [May 5th, 2025]
- WATCH: Journalist Kara Swisher on Elon Musk, whats next in tech and AI | 2025 Reframe Festival - PBS - May 5th, 2025 [May 5th, 2025]
- How an AI Star Wars image has backfired on Trump and the White House - Euronews.com - May 5th, 2025 [May 5th, 2025]
- The AI Industry Has a Huge Problem: the Smarter Its AI Gets, the More It's Hallucinating - futurism.com - May 5th, 2025 [May 5th, 2025]
- I have trouble focusing, but this AI browser feature helps - Fast Company - May 5th, 2025 [May 5th, 2025]
- Its Time To Get Concerned As More Companies Replace Workers With AI - Forbes - May 5th, 2025 [May 5th, 2025]
- Anthropic hires a top Biden official to lead its new AI for social good team (exclusive) - Fast Company - May 5th, 2025 [May 5th, 2025]
- Trump defends viral AI picture of him as the pope: Have to have a little fun - New York Post - May 5th, 2025 [May 5th, 2025]
- Anduril is working on the difficult AI-related task of real-time edge computing - TechCrunch - May 5th, 2025 [May 5th, 2025]
- Datadog Acquires Eppo to Expand Its AI, Product Analytics, Experimentation and Feature Flag Capabilities - GlobeNewswire - May 5th, 2025 [May 5th, 2025]
- Tariffs And AI Are Causing Major Shifts In The Ad Tech M&A Landscape - AdExchanger - May 5th, 2025 [May 5th, 2025]
- AI systems are built on English but not the kind most of the world speaks - The Conversation - May 5th, 2025 [May 5th, 2025]
- Man pleads guilty to using malicious AI software to hack Disney employee - Ars Technica - May 5th, 2025 [May 5th, 2025]
- How to Prevent AI Agents From Becoming the Bad Guys - Dark Reading - May 5th, 2025 [May 5th, 2025]
- This is the future of AI, according to Nvidia - Fast Company - May 5th, 2025 [May 5th, 2025]
- Palantir raises annual revenue forecast on booming AI demand - Yahoo Finance - May 5th, 2025 [May 5th, 2025]
- New ways to interact with information in AI Mode - Google Blog - May 5th, 2025 [May 5th, 2025]
- Google Is Adding Gemini AI to Your Kid's Account, but You Can Turn It Off - Lifehacker - May 5th, 2025 [May 5th, 2025]
- SAG-AFTRA Chief Lays Out What AI Protections It Will Be Looking For In Next Studio Contract - Deadline - May 5th, 2025 [May 5th, 2025]
- Johnson Controls rethinks IT for the cloud-native and AI era - cio.com - May 5th, 2025 [May 5th, 2025]
- AI bots are filling users with conspiracy theories, repressed memories - New York Post - May 5th, 2025 [May 5th, 2025]
- IBM Is Back. Now It Must Prove Its Mettle in AI. - WSJ - April 25th, 2025 [April 25th, 2025]
- Googles AI Overviews now reach more than 1.5 billion people every month - The Verge - April 25th, 2025 [April 25th, 2025]
- Alphabet rises as AI bets begin to pay off - Reuters - April 25th, 2025 [April 25th, 2025]
- Microsoft made an ad with generative AI and nobody noticed - The Verge - April 25th, 2025 [April 25th, 2025]
- Apple to Strip Secret Robotics Unit From AI Chief Weeks After Moving Siri - Bloomberg.com - April 25th, 2025 [April 25th, 2025]
- State Bar of California admits it used AI to develop exam questions, triggering new furor - Los Angeles Times - April 25th, 2025 [April 25th, 2025]
- Heres How Big the AI Revolution Really Is, in Four Charts - WSJ - April 25th, 2025 [April 25th, 2025]
- Update: Meta AI Begins Roll Out on Ray-Ban Meta Glasses to Even More Countries in the EU - Meta | Social Metaverse Company - April 25th, 2025 [April 25th, 2025]
- Adobe Revolutionizes AI-Assisted Creativity with Firefly, the All-In-One Home for AI Content Creation, with New Partner and Firefly Models - Adobe... - April 25th, 2025 [April 25th, 2025]
- Unveiling GPT-image-1: Rising to new heights with image generation in Azure AI Foundry - Microsoft Azure - April 25th, 2025 [April 25th, 2025]
- AI Is Spreading Old Stereotypes to New Languages and Cultures - WIRED - April 25th, 2025 [April 25th, 2025]
- In the age of AI, we must protect human creativity as a natural resource - Ars Technica - April 25th, 2025 [April 25th, 2025]
- Spotify Expands AI Playlist in Beta to Premium Listeners in 40+ New Markets - Spotify For the Record - April 25th, 2025 [April 25th, 2025]
- Microsoft says everyone will be a boss in the future of AI employees - The Guardian - April 25th, 2025 [April 25th, 2025]
- Student loans are back, US travel is whack, and, AI, please, step back : The Indicator from Planet Money - NPR - April 25th, 2025 [April 25th, 2025]
- How real-world businesses are transforming with AI with 261 new stories - The Official Microsoft Blog - April 25th, 2025 [April 25th, 2025]
- This Texas mom made $8,000 in 3 weeks training AI at her kitchen table. She says it's 'not easy money.' - Business Insider - April 25th, 2025 [April 25th, 2025]
- Dataminr Announces $100M Investment from Fortress to Accelerate Gen AI and Agentic AI Product Innovation, and to Expand its Reach to Enterprises &... - April 25th, 2025 [April 25th, 2025]
- Pony.ai teams up with Tencent for robotaxi services on WeChat, other apps - CNBC - April 25th, 2025 [April 25th, 2025]
- Alarming rise in AI-powered scams: Microsoft reveals $4 Billion in thwarted fraud - AI News - April 25th, 2025 [April 25th, 2025]
- CalArts, Chanel Launch Center for Artists and Tech With AI Focus - Variety - April 25th, 2025 [April 25th, 2025]
- China isnt trying to win the AI race - Financial Times - April 25th, 2025 [April 25th, 2025]
- WhatsApp defends 'optional' AI tool that cannot be turned off - BBC - April 25th, 2025 [April 25th, 2025]
- Nvidia Thinks It Has a Better Way of Building AI Agents - WSJ - April 25th, 2025 [April 25th, 2025]
- AI was used to write the California bar exam. The law community is outraged. - Mashable - April 25th, 2025 [April 25th, 2025]
- Exclusive: Anthropic warns fully AI employees are a year away - Axios - April 25th, 2025 [April 25th, 2025]
- Should You Forget Nvidia and Buy These 2 Millionaire-Maker AI Stocks Instead? - The Motley Fool - April 25th, 2025 [April 25th, 2025]
- Opinion: Art is a form of communication between human beings. AI wont change that - The Globe and Mail - April 25th, 2025 [April 25th, 2025]
- Adobe Firefly: The next evolution of creative AI is here - Adobe - April 25th, 2025 [April 25th, 2025]
- Adobe to launch mobile app for AI image generation tool as OpenAI steps up rivalry - CNBC - April 25th, 2025 [April 25th, 2025]
- Humanoid workers and surveillance buggies: embodied AI is reshaping daily life in China - The Guardian - April 21st, 2025 [April 21st, 2025]
- TSMC Warns of Limits of Ability to Keep Its AI Chips From China - Bloomberg.com - April 21st, 2025 [April 21st, 2025]
- A customer support AI went rogueand its a warning for every company considering replacing workers with automation - Fortune - April 21st, 2025 [April 21st, 2025]
- Could AI text alerts help save snow leopards from extinction? - BBC - April 21st, 2025 [April 21st, 2025]
- The #1 Skill That Pays More Than Gen AI In 2025 - Forbes - April 21st, 2025 [April 21st, 2025]
- 1 Artificial Intelligence (AI) Stock-Buyback Stock to Buy Hand Over Fist During the Nasdaq Sell-Off - Yahoo Finance - April 21st, 2025 [April 21st, 2025]
- What America Gets Wrong About the AI Race - Foreign Affairs - April 21st, 2025 [April 21st, 2025]
- Use AI as a tool for growth instead of degradation with this strategy. - Psychology Today - April 21st, 2025 [April 21st, 2025]
- Investor Says AI Is Already "Fully Replacing People" - futurism.com - April 21st, 2025 [April 21st, 2025]
- The philosophers machine: my conversation with Peter Singers AI chatbot - The Guardian - April 21st, 2025 [April 21st, 2025]
- With AI slop distorting our reality, the world is sleepwalking into disaster | Nesrine Malik - The Guardian - April 21st, 2025 [April 21st, 2025]
- Viral AI-made art trends are making artists even more worried about their futures - NBC News - April 21st, 2025 [April 21st, 2025]
- OpenAIs o3 AI model scores lower on a benchmark than the company initially implied - TechCrunch - April 21st, 2025 [April 21st, 2025]
- Artists push back against Barbie-like AI dolls with their own creations - BBC - April 21st, 2025 [April 21st, 2025]
- If you use AI to write me that note, dont expect me to read it - Fast Company - April 21st, 2025 [April 21st, 2025]
- Companies can leverage the true value of meetings with AI by building an LLM for Leadership - GeekWire - April 21st, 2025 [April 21st, 2025]
- Using tech, AI to make construction jobs appeal to women - DW - April 21st, 2025 [April 21st, 2025]
- Famed AI researcher launches controversial startup to replace all human workers everywhere - TechCrunch - April 21st, 2025 [April 21st, 2025]
- Impersonal assistant: This vehicle AI drove me to distraction - Detroit Free Press - April 21st, 2025 [April 21st, 2025]
- A 30-year-old AI founder who followed the FIRE movement to build wealth is now the youngest self-made woman billionaire - Fortune - April 21st, 2025 [April 21st, 2025]