Compliance Training That Passes the Audit and Teaches Nothing

California’s duty sits in Government Code section 12950.1. I set out the counting rules and the records in California harassment training requirements, so here is only the shape of it. The law asks for:

• Seat time: two hours for supervisors and one hour for other employees
• A cycle: every two years, and within six months of hire or promotion
• Required content, including abusive conduct prevention and harassment based on gender identity, gender expression and sexual orientation
• An interactive format, with questions, activities and scenarios
• Records kept for two years

Every item on that list can be checked from a file. None of it can be checked from a hallway.

Someone will eventually ask for proof: a regulator, a plaintiff’s lawyer or a client’s vendor questionnaire. A complete file answers them. The sign-in sheet shows who attended, the certificate shows the hours, and the materials show the topics. That is a real accomplishment, and I would never tell an employer to skip it.

But the file shows that training happened. It does not show that a supervisor could use it. The statute says as much. Subdivision (c) provides that compliance does not insulate an employer from liability for sexual harassment, and subdivision (e) calls the training a minimum threshold. The law that creates the paper trail tells you the paper trail is not the goal.

A hypothetical. A 30-person company buys a one-hour recorded module for staff and a two-hour module for its three supervisors. Everyone finishes, the certificates are saved, and a reminder is set for two years out. Every legal box is checked.

Three months later a supervisor sees a remark about a coworker in a team chat and cannot tell whether it is a big deal. The module covered what harassment is. It did not cover what to do with that uncertainty: who to tell, what to say to the person, what not to promise. The supervisor waits. The file still shows nothing wrong, and nothing was learned.

The most useful public document I know is the June 2016 report of the EEOC Select Task Force on the Study of Harassment in the Workplace. Its argument, in my words, is that much of the training of the previous thirty years was built to limit legal exposure and failed as prevention. It says effective training can reduce harassment, that one design does not fit every workplace, and that training cannot work in a vacuum without a culture and leadership behind it.

Two of its points matter most for a supervisor session. First, the task force called supervisors an employer’s most valuable resource for preventing and stopping harassment and urged employers to train them to respond before conduct becomes legally actionable. Second, it was careful about newer approaches. It saw promise in bystander intervention and civility training and asked researchers to evaluate them. I read that as promising, not proven.

Here are three questions I would put to any session, whatever the vendor or the format:

• Could a supervisor say the first sentence out loud? The first 30 seconds of a disclosure decide most of what follows, as I argued in the manager who freezes
• Do the scenarios look like our workplace, our policy and our reporting path, or like a stock video?
• Does anyone leave with a date and a named next step?

A session can pass every legal test and fail all three. A generic video can cover the required topics. It cannot cover the policy your people actually have or the person they would actually go to. The task force’s point that one size does not fit all is not a criticism of any vendor. It is a description of why a purchased minimum leaves a gap.

That is true, and for a small employer with a tight budget it is a fair position. I am not arguing that every employer needs a custom program. I am arguing that there are two ways to buy the same minimum, and they leave different things behind.

One leaves a certificate. The other leaves a certificate and a supervisor who has said the first sentence out loud and heard how it landed. The hours on the record are identical. Only one of them answers the question the file cannot: what happens in the room when someone finally speaks up.

Keep the statutory hours. They meet the floor. Change what happens inside them. For supervisors, give a share of the two hours to practice: a hypothetical supervisor hears a worrying comment about a coworker, says the first sentence aloud to a partner, gets feedback, and says it again. Use scenarios drawn from the employer’s own settings, with real names removed.

Then measure past completion. Thirty days after the session, ask supervisors what they did with the last disclosure they received, and track whether reports reach HR the same day. I would not promise a result from any of it, and neither can anyone selling you one. These checks tell you whether the training reached the room. For the longer argument on building to more than the floor, see compliance training that does more than comply.

If the budget is tight, the Civil Rights Department publishes free online courses that it says satisfy the state requirement, and using one is a legitimate choice. Then spend whatever time and money you do have on the part no course can supply: a live conversation where supervisors practice what to say.

A passing file is a floor. Treat completion as the start of the measuring, and the audit becomes the least interesting thing about your training.

If you want the practice session run for your supervisors, see the live programs and the OFS page for HR teams. Either way, put a date on the calendar for the 30-day check before the training happens, because that is the step that gets dropped.

Want the supervisor session built around practice?

Bring your current training file. In a free consult we mark what it proves and what it leaves out, then plan the practice.