Congress considers AI and artists’ intellectual property

It’s been a minute since Tech Brew last compiled notes on tech policy news, but we’re back, baby!

In this edition, we’re highlighting cyber and clean energy initiatives from the Oval Office, plus the Senate Judiciary Committee’s ongoing focus on the challenging issue of intellectual property in the AI era.

Here’s what’s been happening at the intersection of technology and politics:

Cyber updates: After releasing its National Cybersecurity Strategy in March, the Biden administration last week published a “roadmap” for its implementation, which includes more than 65 initiatives led by 18 agencies.

Specific initiatives detailed by the administration include an “International Cyberspace and Digital Policy Strategy” to be published by the State Department, and updating the National Cyber Incident Response Plan to include “clear guidance to external partners on the roles and capabilities of federal agencies in incident response and recovery.”

Clean energy campaign: Vice President Kamala Harris wrapped up the administration’s second “Investing in America” tour in Baltimore last week, where she announced a $20 billion investment in clean energy, aimed at “mobilizing a national-scale clean energy financing network.”

The National Clean Investment Fund competition, which is slated to receive $14 billion, will award funds to “national clean financing institutions” to help them provide “accessible, affordable financing” for clean technology projects. The White House said 40% of the funds will go to low-income and disadvantaged communities.

The other $6 billion will give grants to up to seven nonprofits providing “funding and technical assistance” to lenders financing clean tech projects in disadvantaged communities.

(More) AI IP: Senate Judiciary Subcommittee on Intellectual Property last week held its second hearing on AI and IP law.

Artist Karla Ortiz, who said her work was “stolen and used to train for-profit technologies,”

asked Congress to require AI companies to “obtain consent, give credit, pay compensation, and be transparent.”

Stability AI head of public policy Ben Brooks dodged questions about whether owners of copyrighted work know their work is used to train Stability’s products and whether Stability pays owners for that use. Instead, Brooks pointed to Stability’s “opt out” option, which allows creators to request that their images be excluded from training data. (Stability has received opt-out requests for over 160 million images, Brooks said.)

Emory University law professor Matthew Sag told the committee that computer systems should not be recognized as authors under copyright law.

“The Copyright Act rightly reserves copyright for original works of authorship,” Sag explained. “AI can’t produce a work that reflects its own original intellectual conception because it has none.”

“Copying” only takes place in AI models when training data is collected and pre-processed, he added. That’s where companies should adopt best practices to reduce risk of infringement.

“Generative AI does not, in my opinion, require a major overhaul of the US copyright system at this time,” Sag concluded. “If Congress is considering new legislation in relation to AI and copyright, that legislation should be targeted at clarifying the application of existing Fair Use jurisprudence.”