"First, State-by-State regulation by definition creates a patchwork of 50 different regulatory regimes that makes compliance more challenging, particularly for start-ups, reads the order. Second, State laws are increasingly responsible for requiring entities to embed ideological bias within models. For example, a new Colorado law banning 'algorithmic discrimination' may even force AI models to produce false results in order to avoid a 'differential treatment or impact' on protected groups."
"As with any technological disruption, the balance between commercial innovation and public safety is a nuanced and complex area, particularly when it comes to the growth of 'human-like intelligences' as we've seen in the past years of AI research. Innovators need the wide scale commercial protection of consistent, or at least translatable, regulatory regimes to grow into new and existing markets."
The executive order states the United States is in a race with adversaries for AI dominance and that cumbersome regulation, particularly state-level rules, impedes progress. The order identifies three state-level problems: a patchwork of 50 regulatory regimes that complicates compliance for firms and startups; state laws that may require ideological bias in models (citing Colorado's 'algorithmic discrimination' law as a possible example); and state laws that improperly reach beyond borders and burden interstate commerce. The order directs a single federal AI standard that will preempt conflicting state regulations. A Black Duck R&D manager argues that consistent federal guardrails would support innovation while helping balance public safety.
Read at Securitymagazine
Unable to calculate read time
Collection
[
|
...
]