Latest / Roast the Post Podcast / Connecticut AI Employment Law: Disclosure Without Enforcement Teeth? | Grounds for Discussion

Transcript

Connecticut AI Employment Law: Disclosure Without Enforcement Teeth? | Grounds for Discussion · Roast the Post Podcast

• 0:07
  SPEAKER_01: Hey Jen, so I was kind of going through the internet as I do in the morning while
• 0:13
  SPEAKER_01: I'm having my coffee.
• 0:14
  SPEAKER_01: And I found a really interesting headline for us to look at.
• 0:18
  SPEAKER_01: This comes from a law firm in Connecticut, Epstein Becker Green.
• 0:23
  SPEAKER_01: And the headline of what they wrote reads as the following "Connecticut enacts AI
• 0:30
  SPEAKER_01: employment law
• 0:31
  SPEAKER_01: with October 1st compliance deadline."
• 0:36
  SPEAKER_01: For a little bit of context, SB5, which is legislation in Connecticut, was signed
• 0:42
  SPEAKER_01: 27th of May 2026.
• 0:44
  SPEAKER_01: It requires employers to disclose whether AI contributed to mass layoffs starting
• 0:50
  SPEAKER_01: October first, 2026, and it prohibits using automated employment related decision
• 0:56
  SPEAKER_01: technology as a defence in discrimination cases.
• 1:02
  SPEAKER_01: That brings to mind a few questions for me.
• 1:05
  SPEAKER_01: What do you think will be the net impact of these protections?
• 1:09
  SPEAKER_01: Firstly, what do you think this is going to do to the landscape?
• 1:13
  SPEAKER_01: And are there any unintended consequences that might arise from them?
• 1:17
  SPEAKER_01: Because, like, legislation is great until it runs up against the first thing that
• 1:22
  SPEAKER_01: nobody could have foreseen.
• 1:25
  SPEAKER_00: Yeah, I think what might end up happening, and we see this often, is that
• 1:30
  SPEAKER_00: the first hit takes place.
• 1:32
  SPEAKER_00: Maybe there are fines or consequences, companies are penalized or whatever.
• 1:37
  SPEAKER_00: and then what are they gonna do?
• 1:39
  SPEAKER_00: They're gonna find workarounds and they're gonna figure out ways
• 1:42
  SPEAKER_00: to still do what they were doing without getting caught.
• 1:46
  SPEAKER_00: In my opinion.
• 1:49
  SPEAKER_00: So whatever whatever it is the outcome that they are trying to achieve by doing
• 1:53
  SPEAKER_00: what they did.
• 1:54
  SPEAKER_00: They're still going to figure out ways to do it.
• 1:55
  SPEAKER_00: It's just going to look different.
• 1:58
  SPEAKER_00: and then, you know, the legislation will kind of play catch up because every
• 2:03
  SPEAKER_00: policy
• 2:04
  SPEAKER_00: and law and rule, whatever that exists, exists for a reason because they're trying
• 2:09
  SPEAKER_00: to
• 2:10
  SPEAKER_00: correct behaviour.
• 2:12
  SPEAKER_00: And so, so I think that that's what's gonna happen.
• 2:16
  SPEAKER_00: you know, is this going to solve for that problem?
• 2:21
  SPEAKER_00: Yes, to a certain extent.
• 2:23
  SPEAKER_00: and some companies will adhere to it, but there will be workarounds absolutely, and
• 2:31
  SPEAKER_00: then we will see the consequences of that play out.
• 2:34
  SPEAKER_01: Yeah.
• 2:35
  SPEAKER_01: And I wonder too, how do you actually police some of this stuff?
• 2:40
  SPEAKER_01: Right?
• 2:41
  SPEAKER_01: If you're saying, okay, we want employers to disclose whether or not job
• 2:46
  SPEAKER_01: displacement by AI contributed to mass layoffs, how do you how do you police that?
• 2:53
  SPEAKER_01: Let's say that a company does have a mass layoff in Connecticut and a big
• 2:59
  SPEAKER_01: part
• 3:02
  SPEAKER_01: of it is like job displacement by AI, process improvements, technology, whatever
• 3:05
  SPEAKER_01: else.
• 3:07
  SPEAKER_01: but the company knows that this legislation is a thing, and they simply say, no,
• 3:14
  SPEAKER_01: it's because these positions are no longer required or redundant.
• 3:18
  SPEAKER_01: how do you police it?
• 3:19
  SPEAKER_01: Is there a reporting mechanism of some sort?
• 3:22
  SPEAKER_01: This is where legislation that is well intended often kind of comes across comes up
• 3:27
  SPEAKER_01: against the the operational realities of making sure that people
• 3:31
  SPEAKER_01: follow the
• 3:31
  SPEAKER_01: law.
• 3:32
  SPEAKER_01: And I would wonder too about the other part where it said, you know, you can't use
• 3:37
  SPEAKER_01: it as a
• 3:39
  SPEAKER_01: defence that some AI system said fire this person, and you didn't put any agency
• 3:44
  SPEAKER_01: into
• 3:45
  SPEAKER_01: that.
• 3:46
  SPEAKER_01: I would wonder how that is also brought up because none of this stuff is... exists in a
• 3:52
  SPEAKER_01: reality where there's some sort of automated reporting regime where it's just
• 3:57
  SPEAKER_01: magically known.
• 3:58
  SPEAKER_01: by the relevant authorities that something has happened.
• 4:01
  SPEAKER_01: Someone has to know that something bad happened and that it contradicts this
• 4:05
  SPEAKER_01: legislation
• 4:06
  SPEAKER_01: so that they can bring up a complaint or a suit and then it goes through kind of
• 4:10
  SPEAKER_01: the legal
• 4:10
  SPEAKER_01: process.
• 4:12
  SPEAKER_01: All of that is very, very, very, very messy and fuzzy and it'll take a long
• 4:17
  SPEAKER_01: time for it to sort itself out.
• 4:19
  SPEAKER_01: But you have to start somewhere, right?
• 4:21
  SPEAKER_01: So maybe this is a step in the right direction.
• 4:24
  SPEAKER_00: And it is, like at... not that I agree with what I'm about to say, but it's also a very
• 4:32
  SPEAKER_00: grand statement to say that you can't lay off people when a new technology has the
• 4:40
  SPEAKER_00: ability
• 4:40
  SPEAKER_00: to replace that work.
• 4:45
  SPEAKER_00: Again, I don't agree with AI replacing jobs, but at the end the day, a
• 4:49
  SPEAKER_00: legislation gets to say there's this new technology.
• 4:53
  SPEAKER_00: And it's evolving very quickly and it can do all of these things, unprecedented,
• 4:57
  SPEAKER_00: blah, blah, blah.
• 4:58
  SPEAKER_00: But you can't use it to improve your revenue, efficiency, productivity,
• 5:04
  SPEAKER_00: blah, blah, blah.
• 5:05
  SPEAKER_00: you have to...
• 5:05
  SPEAKER_00: I get it.
• 5:07
  SPEAKER_00: I love human beings.
• 5:10
  SPEAKER_00: But it's interesting to me that that's a thing.
• 5:15
  SPEAKER_00: On the other hand, using the AI to determine whether or not somebody can
• 5:20
  SPEAKER_00: keep their job, that's...
• 5:22
  SPEAKER_00: That's...
• 5:22
  SPEAKER_00: I see that a little bit differently because you're using it in a way that
• 5:28
  SPEAKER_00: one is
• 5:30
  SPEAKER_00: about productivity and efficiency and one is about judgment and bias and and that
• 5:35
  SPEAKER_00: kind of
• 5:35
  SPEAKER_00: thing.
• 5:36
  SPEAKER_00: So I see them as two very different uses of AI.
• 5:41
  SPEAKER_01: Yeah.
• 5:42
  SPEAKER_01: Yeah, I mean imagine the printing press, right?
• 5:44
  SPEAKER_01: So the printing press kind of hits this landscape in the sixteenth century or
• 5:49
  SPEAKER_01: whatever,
• 5:50
  SPEAKER_01: and all of a sudden the legions and legions and legions of scribes who were
• 5:56
  SPEAKER_01: typically
• 5:57
  SPEAKER_01: in monasteries, churches, temples or whatever, copying books from, you know, one
• 6:02
  SPEAKER_01: book to a new book with papyrus and a quill and whatever...
• 6:06
  SPEAKER_01: printing press comes around and all those people are redundant all of a sudden.
• 6:10
  SPEAKER_01: Should we have said, no, you can't have the printing press?
• 6:13
  SPEAKER_01: What would society look like today?
• 6:15
  SPEAKER_01: Maybe this will be the same sort of thing in decades, a century, whatever.
• 6:19
  SPEAKER_01: We'll look back on it with the same lens.
• 6:21
  SPEAKER_01: The issue for me, anyway, from a just transition standpoint, there
• 6:26
  SPEAKER_01: are a lot
• 6:26
  SPEAKER_01: of just transitions that are going on in society right now, is, how do you take care
• 6:32
  SPEAKER_01: of
• 6:32
  SPEAKER_01: people during periods of frothiness where there's change and whatnot?
• 6:37
  SPEAKER_01: And that is an issue that
• 6:39
  SPEAKER_01: nobody in government really seems to be paying a whole lot of attention to or
• 6:43
  SPEAKER_01: putting
• 6:43
  SPEAKER_01: their energy to, or maybe just the conversation isn't going in that direction
• 6:47
  SPEAKER_01: because of
• 6:48
  SPEAKER_01: all the other stuff that's going on.
• 6:50
  SPEAKER_01: Who knows?