In a recent workshop on self-editing, I mentioned how Artificial Intelligence could help streamline the process. Once I let the AI cat out of its bag, there was no putting it back.
The conversation drifted to a discussion of Artificial Intelligence and the question of whether it wasn’t more hellhound that devoured your carefully constructed prose, chewed it up, and spit it out after using it as fodder for its training than writing assistant. One of the participants suggested that we read the terms and conditions page of our AI of choice.
Thoroughly rattled by what I heard (What do you mean that the AI can claim my work as its own? I’ve already input a chunk of my work into a several AIs—what have I done?) I went home and did just that. Here’s what I found.
ChatGPT
I did a lot of brainstorming with ChatGPT. I’ve also asked it to critique several chapters of my book to see what it would say.
According to OpenAI’s terms and conditions, “you (a)retain your ownership and rights in Input and (b) own the Output.” They go on to say that “We hereby assign to you all our right, title, and interest, if any, in and to Output,” which is good news. ChatGPT cannot claim any of my work as its own, nor can it claim anything the AI generates for me as its own.
Whew!
I can also opt out of allowing them to use my content to train their model, which I did before pasting word one of my text into the interface.
ChatGPT seems safe, but don’t take my word for it. If you are going to use ChatGPT as a writing assistant, read the terms and conditions and decide for yourself.
Google Gemini
When ChatGPT was lazy or gave me sass, I switched to Google Gemini. I found it the better AI for critiquing my work, but I also used it for brainstorming and to summarize articles for research when world-building.
Google’s Privacy & Terms document says that it “will not claim ownership over [your] content.” However, you do give Google permission to use your content if it is made public and available to others or to “modify and create derivative works based on your content, such as reformatting or translating it” with the purpose of “operating and improving the services,” including training Gemini, Google’s AI. If you are worried, you can opt out of allowing Google to use your content to train its AI. To learn more (and I do encourage you to learn more) read the complete Privacy & Terms document before using Gemini, especially as a writing assistant.
Another sigh of relief.
Microsoft Copilot
I’ve never tried using Copilot as a writing assistant, but I know people who say it does marvelous things, so I decided to check it out.
Microsoft’s terms and conditions for Copilot state that they “don’t own Your Content, but [they] may use Your Content to operate Copilot and improve it. By using Copilot, you grant [them] permission to use Your Content, which means [they] can copy, distribute, transmit, publicly display, publicly perform, edit, translate, and reformat it, and [they] can give those same rights to others who work on [their] behalf.”
It goes on to say that they “get to decide whether to use Your Content, and [they] don’t have to pay you, ask your permission, or tell you when [they] do.”
Pretty scary, huh?
The webpage refers you to a second document that smooths this over a bit by saying that it will only use your Copilot conversations to improve the AI or personalize your conversation. You can opt out of the training and personalization, but giving Microsoft the ability to use my content as it sees fit, without notifying me or asking permission first, is essentially granting it joint copyright. That’s not something I’m willing to do after pouring years of blood, sweat, and tears into my manuscript. The decision is ultimately up to you, but be sure to read the terms and conditions before you make it. Moreover, Artificial Intelligence systems each have different policies so users should check them carefully.
Conclusion
AI is ever-advancing, and new interfaces are popping up at an alarming rate. Artificial Intelligence continues to grow and change rapidly. Using AI is a catch-22 in many ways. It can be a huge help for creators, but every bit of information we enter is used to teach it how to be more “human.” Currently, AI has difficulty understanding human emotions and the way things work in the real world. Flashforward two, five, or ten years of people inputting information indicative of the human condition—will that eventually change?
For now, all we can do is protect human-generated work from being consumed and repackaged by AI. Know how your AI of choice treats and stores your content. Know who owns your input and output. You wouldn’t trust any old person as your writing assistant without interviewing and fully vetting them. Do the same when looking for an AI writing assistant, and read the terms and conditions of the software. In summary, responsible use of Artificial Intelligence requires vigilance and care.
Leave a Reply