SEC: AI needs Machine Readable Data
At XBRL International we’ve been making the point for a long time that to take advantage of AI and other advanced analytic tools, you need high quality, digital, machine readable data and metadata. In a speech to the Financial Information Management Conference yesterday, Scott W. Bauguess, Deputy Chief Economist and Deputy Director, Division of Economic and Risk Analysis at the SEC, described a regulator’s viewpoint on this subject.
Acknowledging the importance of (increasingly open source) AI algorithms, and the right people to use them, Bauguess emphasises the critical part that machine readable data plays. The massive EDGAR database serves up 1.5billion copies of documents filed by public companies with the SEC each year, and these days 85% of the downloads made each day are made by ‘bots, not humans.
As he says, the “SEC is fundamentally committed to ensuring that all investors and market participants can access the information necessary to make informed financial decisions”. Unsurprisingly an increasing number of participants want to use advanced technology like AI to help them make those decisions, and those “bots” downloads are a big hint!
To overcome this, the Commission is looking to ensure that humans and machines can read their reports equally well, and, through a proposed rule, is seeking to use Inline XBRL to ensure this happens. Mr Bauguess points out that “[t]his proposed rule, if adopted, would combine the two requirements and create a single document designed to be read equally well by humans and machines.”
The SEC is concerned not just with market access to this information, it also wants to use it themselves, primarily to discover misconduct. It already uses an array of AI tools to examine the data it has at its disposal and is seeking to improve its capabilities by improving the usability of the data it receives.
Read about the Five Myths about Machine Readable Data Standards that he describes. They are very much worth your time. In fact, this speech should be mandatory reading for anyone interested in the future of reporting. It makes plain that it is digital and at the same time human and machine readable. That’s Inline XBRLto you and me.