BlastPoint

A model making the decision doesn't make fair lending rules go away.

If you're using AI or other models to inform lending decisions, you still need to understand how those decisions are being made and whether they're treating people fairly.

The technology changes. The responsibility doesn't.

[LIVE CU ROUNDTABLE] What happens after a member opens their first product?

For many credit unions, that’s where the relationship starts to stall.

On October 22, Tomer Borenstein will host a conversation with Steven Reed from P1FCU about how they approached getting more members into a second product, what they learned, and what they’d do differently.

We’ll also hear from other credit union leaders about what they’re seeing, what they’ve tried, and how they’re thinking about deepening member relationships.

Request a seat: https://hubs.li/Q04yZtTy0

A few seats are still open for our Oct. 7 conversation with Single Stop!

What happens when you can spot a customer heading toward trouble before they fall into arrears?

That’s what we’re getting into with Single Stop and Hillary Wi******er from Seattle City Light.

We’ll talk about what utilities are doing to find these customers, connect them with help, and actually know whether it worked.

If affordability and arrears are on your list right now, come join the conversation.
https://hubs.li/Q04yYlcY0

“The model said so” isn’t an explanation.

When a model makes a decision, you should be able to understand why.

What data influenced the outcome?
What factors mattered?
Can you trace the decision back to something you can actually evaluate?

And just as important: can you audit that decision for bias?

As more financial institutions use models to inform decisions, both matter. Because you can't meaningfully question, evaluate, or improve a decision you can't explain.

The goal isn't to make algorithms less powerful. It's to make their decisions more transparent, explainable, and auditable.

A closer look at energy affordability, from the customer perspective.

Chris Medina joins SECC and other industry leaders this morning for the SECC Workshop at GridFWD, sharing findings from nationwide customer research.
10 a.m.– 12 p.m.

If you’re at GridFWD, come hear what customers are saying and find Chris and Dave Graswick while you’re there.

A few spots left for our Oct. 7 utility roundtable with Single Stop.

We’ll be joined by Hillary Wi******er from Seattle City Light for a candid conversation about what happens when a customer is at risk of falling behind, but hasn’t fallen into arrears yet.

We’ll talk about how utilities can:
• Identify customers earlier
• Connect them with assistance they may already qualify for
• Track what happens next

If you’re working on affordability, arrears, or customer assistance, there’s still time to join us. https://hubs.li/Q04yDbTP0

Before you hand your data over to an AI vendor, it’s worth knowing what you should be able to expect from them.

We put together 14 rights every organization can expect from any AI vendor it works with — starting with where your data lives, how it’s kept separate, and what vendors can (and can’t) do with it.

Take a look through the first three.

Want all fourteen? Comment RIGHTS and we’ll send them to you.