In 2011, a service technician at a California auto dealership group signed an arbitration agreement during onboarding. The document included his name, an electronic signature notation, and a timestamp. When he later filed a wage-and-hour class action, the company tried to enforce the arbitration agreement. The employee said he didn’t remember signing it.
In Ruiz v. Moss Bros. Auto Group, the California Court of Appeal upheld the denial of the company’s petition to compel arbitration. The employer could produce the signed document but couldn’t adequately explain how the electronic signature was applied or establish that only the employee could have done so. The court indicated that information about the login process, security procedures, and steps required to accept the agreement could have helped authenticate the signature.
The case illustrates a fairly common problem with electronic agreements. Businesses often retain the final signature or acknowledgment without keeping much information about the process that produced it.
A Signature Records the Outcome, Not the Process
A signed document shows that a signature was placed on a page. A checked box shows that an acknowledgment was recorded. Depending on the system, those records may not show who completed the action, which version of the agreement was displayed, or how the information was presented.
For routine paperwork, that distinction may not matter. It becomes more important with arbitration agreements, data-use permissions, liability waivers, safety policies, auto-renewal terms, and other agreements that may be disputed later. Digital consent can make the gap more noticeable because users can move through agreements very quickly.
Researchers Jonathan Obar and Anne Oeldorf-Hirsch demonstrated this with the NameDrop terms-of-service experiment. Participants were asked to join a fictional social network and were presented with terms and a privacy policy. Most spent little time reading them, and almost no one noticed the deliberately unreasonable provisions included by the researchers. The participants still completed the consent process. From a recordkeeping standpoint, the experiment shows why an acknowledgment alone doesn’t necessarily provide much information about what happened before someone clicked “agree.”
Clinical Research Has a More Detailed Approach
Clinical trials have had to address this issue because informed consent is required to enroll human participants in research. A signed form is important, but research teams may also need to establish that participants received the appropriate information and the correct version of the consent materials. Communicating that information clearly can still be difficult.
An analysis of 5,239 clinical trial consent forms found that many were lengthy and written at reading levels above commonly recommended levels. The findings highlight a basic problem with consent: providing information isn’t particularly useful if the intended audience has difficulty understanding it.
Electronic systems have given research teams more ways to document the process. Platforms for electronic informed consent in clinical trials can track details such as which version of a form was presented and when it was signed. When consent materials change, electronic systems can also help identify participants who need updated information or renewed consent. The useful distinction for businesses is between recording the final action and maintaining enough context to explain how that action occurred.
Businesses Have Similar Gaps
The same basic issue appears in employee, customer, and contractor agreements. Employee onboarding is one example. A new hire may receive a handbook, arbitration agreement, remote-work rules, expense policies, and other documents through the same portal. If a policy becomes relevant months or years later, the company may need more than a record showing that the employee acknowledged “the handbook.” It may need to establish which version was available at the time and how the acknowledgment process worked.
This becomes more important as workplace policies cover newer areas of risk. B2BNN’s discussion of unanswered liability questions around corporate rideshare is one example. Employers can establish rules about when company-sponsored transportation should be used, but they also need a reliable way to distribute and document those policies.
Customer agreements create similar issues. Software companies regularly update terms of service, privacy policies, and data-use practices. AI features can introduce additional changes in how customer information is processed or used. If the company’s record simply shows that a customer agreed to “the terms,” it may not be clear which version of those terms was actually presented.
Internal systems can create another variation of the problem. B2BNN has discussed AI-driven insurance platforms where decisions across disconnected systems can result in audit trails stitched together after the fact. Consent records can have a similar weakness when different pieces of information are stored separately without a clear record of the overall process.
What Should a Consent Record Include?
Businesses don’t need clinical-trial infrastructure for every agreement. The appropriate amount of documentation depends on the importance of the agreement and the consequences if it’s challenged.
For higher-stakes agreements, useful records may include:
- Who accepted the agreement and how identity was authenticated
- Which version of the document was presented
- When the document was presented and accepted
- Whether important terms required a separate acknowledgment
- Whether the person received a copy
- Whether later revisions required another acknowledgment
Many electronic signature, HR, and customer-management platforms already capture some of this information. Businesses may simply need to review how those features are configured and what records are retained.
The process doesn’t have to be equally detailed for everything. A newsletter signup doesn’t require the same documentation as an arbitration agreement. A minor product update may not require the same consent process as a significant change in how customer data is used. The amount of documentation should generally reflect the significance of the agreement.
Clear Language Matters Too
Recordkeeping is only part of the issue. The agreement itself also needs to be understandable. Clinical trial research on lengthy and complex consent forms shows that even heavily regulated consent processes can struggle with readability. Businesses can run into the same problem when important terms are buried inside long policies or written in unnecessarily technical language.
Plain language can make those terms easier to understand. Particularly important provisions can also be presented separately rather than relying on the user to identify them within a much longer document.
Keep the Context With the Signature
In some situations, a brief acknowledgment or comprehension check may also be useful. An employee could separately acknowledge an arbitration provision, for example, or a customer could confirm a significant change in how their information will be used. The goal is simply to create a clearer connection between the important term and the person’s agreement to it.
Electronic signatures and checkboxes aren’t inherently weak forms of consent. The problem arises when they’re the only information a business retains. A stronger record connects the final action to the surrounding process: who completed it, which document they received, when they received it, and how the agreement was presented.
Clinical research provides a useful model because those systems are designed to document more than just the final signature. Businesses don’t need to replicate the entire process, but they can apply the same basic principle to agreements with meaningful legal or financial consequences.
The practical question is what the company would need to establish if an agreement were challenged later. Building that information into the consent process is much easier than trying to reconstruct it afterward.
