Sacramento, Calif. – As California legislators consider several tech-focused bills, including proposals regulating artificial intelligence and reporting mechanisms for cyberbullying, the Computer & Communications Industry Association will testify on Tuesday.
CCIA opposes AB 2655, primarily due to the bill’s reliance on the flawed assumption that online platforms possess the capability to definitively and reliably detect manipulated content as defined by the bill. CCIA has concerns about SB 1504 because, much like the bill focused on labeling and blocking user-generated content deemed to be “materially deceptive”, it targets digital services rather than the bad actors doing the bullying. These approaches, unfortunately, do not effectively target the root cause of the harms the legislation intends to address.
As for AB 3211, which aims to establish requirements for watermarking and digital provenance for AI-generated content, CCIA will testify about the shortcomings of the current technology and flag the risks of legitimizing inauthentic content and eroding trust online.
The following can be attributed to CCIA State Policy Director Khara Boender, who will be testifying on these bills:
“We share legislators’ goal of curbing cyberbullying and the spread of misinformation ahead of an election. In both cases, we would recommend they target the bad actors spreading misinformation and engaging in bullying. Failing to hold those spreading problematic content accountable perpetuates a cycle where online platforms continually contend with such material without proactively preventing its occurrence.”
“California lawmakers are aiming to get ahead of any issues with AI generated content online, but the technology is not at a stage to fit well with their proposed solution of watermarks, and could lead to accidentally authenticating misinformation. It’s worth keeping in mind that faked material being treated as real is (generally) worse than real material being treated as fake.”
