ALLi News Editor, Dan Holloway
A few weeks ago I reported on the launch of Fairly Trained. This is a group that sought to be a trade certification body for AI industries that was able to provide a stamp of approval for the methods and motives of individual companies and platforms in much the same way as the Forest Stewardship Council (FSC) does for the timber industry. And of course its name is a play (as opposed to a plagiarism – wouldn’t that be ironic!) on the most famous such certificating body of all, Fair Trade.
I was sceptical at the time of Fairly Trained’s Launch. Unlike OpenAI, who claimed in 2023 that it would be impossible to train AI without copyright material, I don’t think possibility drove my scepticism. It was more what felt like pragmatism (and may have come off as cynicism) about how markets work. That is, when there are such powerful tools available to so many that have a past mired in copyright mud, would anyone want to use an AI platform that was ethically trained? And would their customers in turn be interested?
In the non-digital world, certification usually comes with a premium cost that is offset by three things on the consumer side:
- The ability to pay a small premium in enough numbers to make the model work for suppliers;
- A desire to do the right thing;
- An understanding of how what is being presented represents the right thing.
I don’t doubt that many of the end users of AI have the second of these qualities. Financial duress has increasingly shown that even in well-established markets, the first is shrinking. My real doubt was over the third. Copyright is complex. Creators are passionate about it, but the public?
Well, there is clearly enough belief that there’s a use case for Fairly Trained to have announced their first certification. It goes to KL3M. What’s really interesting is that the company behind it is 273 Ventures, a legal tech start-up. In other words, this is not a case of serving an audience wanting to do the right thing. It’s about targeting an audience terrified of doing the (legally) wrong thing. And thinking about it like that, it makes a lot of sense. Compliance and risk avoidance are massive spaces, and are only getting more so. It might take time for the public to catch on. But in the meanwhile, this may be a market with enough legs to prove the concept.
