OpenAI has launched Decisions, its own decision system which will compete with the likes of Jev (see here). It could potentially reduce AI costs for bulk doc review and provide other benefits to lawyers.
First, what is Decisions? The short version is that it’s a ‘sorting system’ that uses AI, but in a much more limited way compared to tapping the full resources of an LLM. You can call it ‘AI lite’ perhaps.
Why would lawyers want to use this? The first reason is cost – see tables below. Because these decision approaches are not digging that deeply into the LLMs they use, the token costs are much lower. For example, for now with Decisions (which relies on GPT-6-Luna) the input costs are $0.10 per 1M tokens. You pay only for input tokens: there are no cache-read, cache-write, or output-token charges. That is a big reduction in price.
All well and good, but what can it do? It’s best at simple classification and really is not for complex language understanding or legal reasoning. So, you could feed in 10,000 documents and get it to sort them via a range of categories. It also handles images very well – but that’s less useful to lawyers.
As OpenAI explained: ‘The Decisions API [which is in public beta] evaluates text, images, or both and returns typed answers about 10x faster than the Responses API.
‘Get the probability that a condition is true, a choice from a fixed set, or a score against a rubric. Use those answers to classify content, route requests, and prioritize work in your application. Both choice and score return probabilities over discrete options.’
In short: look for X, Y, Z – are those things there? – mark them as such based on the probability they are those things.
So, this could work well for first-pass review in a due diligence project, or for eDiscovery, or large-scale compliance reviews. Basically anything where lawyers want to get into a large set of documents, and after that they can drill down in more detail using more traditional LLM-based approaches.
What it won’t be good for is things like complex contract review, or any other demand where there needs to be both subtle language understanding and sophisticated outputs. Instead, this approach very much has ‘volume’ written all over it.
AL asked OpenAI’s ChatGPT to build some tables showing what Decisions can do and how it compares to Jev.
Conclusion
The cost savings for use cases such as first-pass review are very significant, thus most large commercial law firms will want to have look at this.
As noted, this is not an alternative to using the main models in the usual way, or the legal AI tools that tap the full capabilities of those models, when it comes to more subtle and sophisticated work. But, if you have a bulk document project then this could be a useful approach that saves a lot of money.
More about OpenAI Decisions here.
And here’s a short video by OpenAI as well.
—
Legal Innovators UK on Nov 4 and 5 in London, has now opened for Express Registration so you can get your tickets super-fast!
If you qualify you can very quickly receive your tickets for the landmark innovation event – follow the links below. Also, our Legal Innovators New York event is now doing the same. We look forward to seeing you on either side of the Atlantic to explore the cutting edge of the latest legal AI developments, what it all means, and the very latest products too!
Legal Innovators UK – Nov 4 and 5 / Legal Innovators New York – Nov 17 and 18
For London – Nov 4 and 5, at the Minster Building, see here:
And, for New York – Nov 17 and 18 – Midtown, please see here:
Discover more from Artificial Lawyer
Subscribe to get the latest posts sent to your email.