Be honest: Do you really read the terms of service before clicking “Agree”? How about a legal document? Do you actually read every word before signing your name?
Thoroughly reading a document and understanding its terms before signing is a vital part of the process of agreeing to terms — or maybe not according to Docusign.
The “Q-Tip” Problem
While they don’t explicitly advertise this as a way to avoid reading, it’s hard to imagine people using it any other way. We may have the AI version of Q-Tips here: they aren’t supposed to go in your ears, but we all know that’s exactly where they end up.
The tool, powered by an AI engine called Iris, generates a “plain-English” summary and allows signers to ask questions. Docusign claims this makes complex agreements easier to understand, allowing people to sign “faster, with confidence.”
Confidence vs. Correctness
“Confidence” is the key word in Docusign’s marketing. Their survey found that 75% of people would feel more confident signing if assisted by AI.
There’s no shortage of LLMs available to consumers, but Docusign believes its new AI-assisted signer experience is better than pasting a contract into a generic AI tool.
That’s important, because AI tools often “hallucinate” responses, meaning they confidently state information that simply isn’t there. Some of those are small gaffes, but failing to fact-check AI-generated responses can have severe consequences.
Recently in England, West Midlands Police revealed that an “erroneous result” in an intelligence report claimed a football match had occurred and had an incident. That report contributed to police barring Maccabi Tel Aviv fans from attending a match against Aston Villa.
In a police report, that’s a potential civil rights issue. In a contract, an AI hallucination could mean missing a predatory “gotcha” clause or a hidden fee.
More data, less problems?
Docusign argues that Iris is different because it isn’t a general-purpose chatbot. The engine uses “contract-specific data” and the company’s IAM (Intelligent Agreement Management) platform to be more accurate.
The Iris AI engine has processed over 100 million customer agreements since last year, creating what Docusign calls a “data advantage.”
The capabilities of any LLM depend largely on the data used for training. For example, Google has access to a lot of images and its image generation tools are ahead of the competition.
Mangesh Bhandarkar, GVP of Product Management at Docusign says the new AI features can “eliminate signer hesitation.”
I just wonder if hesitation is a negative when it comes to signing agreements.
Docusign assumes we’re tired of “contract busywork” and will trust AI to summarize documents. Are you ready to stop reading the fine print and rely on the Iris engine?
Follow Windows Central on Google News to keep our latest news, insights, and features at the top of your feeds!
