Q&A
Discussion on the potential for enforcement in future AI monitoring
0:41:06
·
57 sec
The discussion explores the potential for enforcement mechanisms in future AI monitoring processes. Alex Foard explains that it's too early to specify details about future policies that haven't been drafted yet.
- The goal of monitoring is reflective of a fuller picture of governance
- OTI is working through a sequence of steps to build out the governance framework
- Foard emphasizes the need to work through earlier pieces of the framework before committing to later pieces like enforcement
Jennifer Gutiérrez
0:41:06
And monitoring.
0:41:07
So it's it's something that you're building out, as I understand, would monitoring in the future, at some point, lead to some kind of level of enforcement or monitoring is just we've assessed.
Alex Foard
0:41:19
I I would hesitate to to speak to you specifically about what our future policy that we haven't drafted yet.
0:41:24
I will exactly say, but the goal of monitoring is reflective of that of that fuller picture of governance.
0:41:30
Right?
0:41:30
So the full first section of the action plan is committed to building out governance, and that is is sort of across that life cycle.
0:41:37
Right?
0:41:38
So for you start at the foundation with principal's definitions, you move into how we can provide agencies with preliminary guidance on some of the major technologies that they'll care about.
0:41:47
Then we move into how we understand and and account for potential for risk, and then talk about how we monitor that over time.
0:41:54
So you can see there's a sort of sequence that needs to be constructed from this, and we need to work through the earlier pieces before we before we sort of commit to some of those later pieces.
