Discover more from The Author Analyst
Kickstarter's New AI Policy Goes Into Effect
Some huge clarification and some revelations within the author community
Today is September 1 and there's a new policy in effect at Kickstarter–projects need to comply with the new AI Policy at Kickstarter.
I originally wrote a post about it on Author Analyst, which you can read here:
There are a few further new-to-me developments in both the policy and the chatter amongst the author community that I wanted to share here. While most of this is just a roundup of the current conversation in author communities about the topic, I’ve also added my initial thoughts to each point:
#1 - Kickstarter Clarifies How Authors and Artists Are Responsible For Consent of Data Training AI Tools
Long story short, if you are an author and had issues with the Consent portion of the policy, good news—it only applies to developers and apps, not those who are simply using tools like Midjourney or ChatGPT. If you had issues with the Transparency portion of the policy…Well, not much has changed on that front since the announcement.
If you read my original post, you know that I struggled with understanding how exactly a creator would be responsible for tracing Consent back through a third-party software company—and now we have the answer. We are not expected to do that. And I’m incredibly glad to see that clarified more officially ahead of the implementation.
It’s not everything I took issue with from the policy. I understand that many have very reasonable issues around Transparency still—me too— and I’m hoping those will be addressed over the long-term.
That said, it is a positive step and significantly loosened stance on liability around Consent, which I think is a good thing.
#2 - Kickstarter Launches Additional Questions Several Days Early, and On Projects That Have Already Been Approved But Not Launched
Kickstarter stated that the AI policy would go into effect for projects approved after August 31st.
Around August 30th, a message started cropping up on already approved projects that hadn't launched yet, that the project was incomplete. This had the author community a little up in arms, especially those who had worked to get their projects approved before the August 31st deadline.
It's still not clear to me whether a project that's already approved can continue to launch without filling out this information, but I think we'll know next week as a lot of people in this boat will launch after the holiday weekend here in the US. (Tuesdays are the most popular day to launch, so I’m expecting to hear a few takes by September 5th.)
I don't have a project affected by this right now so I can't test it myself. Until I hear back that people are not able to launch already approved projects without filling in the information, I'm hesitant to say that Kickstarter has gone back on their original description of how this policy will go into effect.
However, even the threat of this not working the way it was stated it would is understandably soured people's experience of the platform and further eroded trust during an already fragile transition.
#3 - As This Policy Goes Into Effect, Authors Are Watching Closely For How It Changes the Platform
There are three things I'm personally watching for now:
Actual algorithmic changes - Since Kickstarter is collecting data on AI projects on their platform more rigorously, it begs the question if they will be updating their discoverability and visibility algorithms to boost some projects or suppress others. The answer is simply that we do not know yet—but we will be able to see it right away, within a few weeks, if it’s happening. Kickstarter’s algorithms are incredibly straightforward and simple to watch. As a side note, people tend to take their personal experiences and extrapolate them to algorithmic changes. These are not things I report on unless I see enough relevant data to back it up and make my own call. I have high standards in this area and seek confirmation rather than speculation. It's very easy for the author community to get up in arms about things that aren't actually happening, and I will do my best to be responsible in reporting, and not fuel the spread of rumors.
Kickstarter giving less visibility to AI projects - There are many manual elements that drive the Kickstarter algorithms, including Projects We Love. Further, there is extra visibility on Kickstarter through merchandising opportunities like Project of the Day (by category). Even if some projects are not algorithmically suppressed, creators could still be harmed by being passed over for these opportunities.
Backer vitriol toward projects created with AI - It's no secret that AI is polarizing. The people who want to know if AI is used are obviously driving this policy forward, and they are also more likely to share the view that AI is morally or ethically wrong and creators shouldn't be using it. I'm personally torn on the topic. On the one hand, no one is entitled to backers, and backers should be able to back what aligns with their beliefs. On the other hand, it's a privilege to be Transparent—it always has been—to the point where countries now have laws to protect nondisclosures of race, gender, and more. Kickstarter has a reputation for being pro-female, pro-sex, pro-LGBTQIA+, but their Transparency policies can sometimes conflict with that mission. It's something I've written to them about multiple times, and something I'll be writing to them about again.
My Final Take At the Moment (Subject to Change, Of Course!)
I’ll be watching the developments in this space, and I’m hoping for more clarification in the coming weeks. I would love to see Kickstarter continue to be a place where authors feel good about spending their time and energy!