ELG Law

California's AI Safety Bill SB 1047 ("Safe and Secure Innovation for Frontier Artificial Intelligence Models Act") was amended last week after passing through the Appropriations Committee. These amendments introduce the following changes: ? the California AG can no longer sue AI companies for negligent safety practices before a catastrophic event has occurred ? rather, the AG can seek injunctive relief that would require the company to cease a certain operation if found to be dangerous, and can still sue an AI developer if its model does cause a catastrophic event ? the law would no longer create the Frontier Model Division (FMD), a new government agency formerly included in the bill ? AI labs no longer need to submit certifications of safety test results "under penalty of perjury"; now, they only need to submit public "statements" outlining their safety practices and the bill imposes *no criminal liability* ? the standard for ensuring model safety has now changed from requiring developers to provide "reasonable assurance" to requiring developers to provide "reasonable care" The amendments come following intense opposition to the bill from industry leaders and entities claiming that the law would stifle AI innovation. Because they still hold developers liable for the dangers of their AI models, opposition to it has not wavered. The law is a significant deviation from the stance that the U.S. has historically taken on AI innovation in cohesion with other countries like Japan and South Korea. It'll be interesting to see whether it passes a final vote. #artificialintelligence #AI

The Illinois Human Rights Act, which governs employment discrimination, has been expanded to include #AI regulation. Last week, Illinois Governor J.B. Pritzker signed legislation that will require Illinois employers to inform workers and job applicants about their use of #artificialintelligence technology in making employment decisions. The new law also prohibits the use of AI that leads to discriminatory practices against employees. The law offers amendments to the Human Rights Act by listing the use of AI, including #generativeAI, as a potential violation of civil rights. The law is set to take effect on January 1, 2026 so employers will have time to assess and prepare for compliance. Employers should: 1?? Assess their AI systems ? assessment should involve analyzing algorithms, data inputs, and outcomes to identify any potential issues 2?? Notify workers ? inform workers and job applicants about the use of AI in employment decisions 3?? Update policies and procedures ? create guidelines for the appropriate use of AI, conduct impact assessments, and implement safeguards to prevent discrimination 4?? Train employees ? train employees involved in recruitment, hiring, promotion, and disciplinary processes on the proper use of AI, including on how to identify and mitigate biases 5?? Monitor compliance ? establish mechanisms to monitor compliance, including regular audits, reviews, and assessments of AI usage in employment decisions

Valuable breakdown of the Copyright Office's first AI report on #deepfakes #AI #artificialintelligence

California's #AI Bill, the Safe and Secure Innovation for Frontier #ArtificialIntelligence Models Act is entering its final stages in the legislative process. The law would require AI companies to train large-scale models to conduct rigorous safety testing for potentially dangerous capabilities and implement comprehensive safety measures to mitigate risks. It would require that a developer, before initially training a covered model, comply with various requirements, including implementing the capability to promptly enact a full shutdown, as defined, and implement a written and separate safety and security protocol, as specified." It would further prohibit a developer from using a covered model commercially or public, or making a covered model or a covered model or its derivative available for commercial or public use, if "there is an unreasonable risk that the covered model or covered model derivative can cause or enable a critical harm, as defined." The bill has faced much opposition from industry and investors who claim it would stifle innovation. Pushback on the bill centers around provisions in the law which would place a burden on developers to provide reasonable assurances that an AI model will not pose unreasonable risk of causing "critical harms" such as aiding in the creation of weapons of mass destruction or causing severe damage to critical infrastructure. The bill would apply to systems that both cost over $100 million dollars to train and are trained using an amount of computing power above a specified threshold. Promoters state that these requirements imply that the bill would likely only affect the largest AI developers. However, detractors claim that the law stifles innovation, that large AI developers have already made voluntary commitments to undertake safety measures, and it would harm the open-source community. The bill does steer away from the U.S.' position on AI regulation, which has tended to be in favor of innovation over regulation. California is leading states in AI governance statutes and this bill could make it an even bigger anomaly.

The U.S. Copyright Office has released the first part in a series of reports addressing the legal and policy issues related to #artificialintelligence and #copyrightlaw. The first part addresses digital replicas, otherwise known as "#deepfakes". The report urges the passage of a federal law to confront the rapid proliferation of deepfakes, analyzing existing legal frameworks governing copyrights. According to the Office, it has received many comments seeking protection against #AI "outputs that imitate the artistic style of a human creator." "The speed, precision, and scale of AI-created digital replicas calls for prompt federal action. Without a robust nationwide remedy, their unauthorized publication and distribution threaten substantial harm not only in the entertainment and political arenas, but also for private individuals." Specifically, the law is calling for a law that is "narrower than and distinct from" existing laws addressing #copyrightinfringement. This is due to the additional layer of implicit replication and the potential for harm. The new copyright law should apply beyond the commercial market, as individuals can harm one another with deepfakes. The law should establish "a federal right that protects all individuals during their lifetimes from the knowing distribution of unauthorized digital replicas. The right should be licensable, subject to guardrails, but not assignable, with effective remedies including monetary damages and injunctive relief." The Office suggests that the law apply only to the "distribution or making available of an unauthorized digital replica, but not the act of creation alone." Creation of an otherwise unauthorized deepfake for personal use could be legitimate part of the creative process. This is the first report in a series that will focus on related topics, including the ability to copyright works created using generative AI training of AI models on copyrighted works (prevalent for the current litigation landscape), licensing considerations, and allocation of any potential liability. Full report here ? https://lnkd.in/gBun9uBA

? TikTok's Legal Woes and #Privacy Investigations Continue The FTC has officially referred its investigation of TikTok to the Department of Justice. The agency stated that its investigation uncovered reason to believe that TikTok is violating or about to violate the FTC Act and the Children's Online Privacy Protection Act (#COPPA). This referral comes soon after Congress' passage of a law requiring divestiture of TikTok by its Chinese parent company and sale of the company to a government-approved buyer. The agency, however, did not specifically articulate its findings. The public statement is also an unusual move as the FTC does not typically publicize its referral decisions. The agency stated that it felt doing so in this instance "was in the public interest". TikTok has come under extensive and continuous fire by Congress, the FTC, and state agencies over the last year or two for its child online safety failures and data protection practices. For example, the company is currently being sued by the State of Utah for allegedly facilitating the exploitation of children via TikTok LIVE. The FTC's announcement states that their investigation began in connection with their "order compliance review of Musical.ly, [TikTok's successor], following a 2019 settlement with the company for violations of the Children's Online Privacy Protection Act ("COPPA")." As child online safety and privacy becomes more of a pressing issue not just in the United States, but globally as well, organizations will have to be more proactive in ensuring compliance with data collection/privacy and safety laws. #dataprivacy #socialmedia #onlinesafety #childprivacy #artificialintelligence https://lnkd.in/giKPpNpE

To learn more about how #deepfakes are challenging the #law, check out our founder Donna Etemadi's article, "The Deepfake Dilemma" published in the June edition of the Illinois Bar Journal! #Deepfakes are a particularly troubling byproduct of #artificialintelligence. Deepfakes have and will continue to present a plethora of social and legal issues for individuals such as comedians, public figures, and school students, and entities such as online platforms and news organizations. They pose a serious danger to youth and in the age of social media, will likely become impossible to contain, having widespread effects on youth mental health and cyberbullying. They also have the capacity to present serious national security and public safety risks. While attempts to regulate deepfakes have arisen both on a state and federal level, additional efforts are necessary to bridge the information gap between policymakers and industry to promote the development of safe and responsible #GenerativeAI #technology and stem the proliferation of harmful deepfakes. ?? Article available here >> https://lnkd.in/gcbch33b

ELG Law is now open! We are a multi-service firm leveraging experience in the public and private sectors to provide counsel on matters ranging from general corporate to regulatory counseling and litigation (including #deepfake litigation and #AI product liability) to public policy advisory services, with a focus on #technology, #dataprivacy, and #artificialintelligence. We provide services in California, Chicago, and Washington D.C. Data privacy continues to be a pressing social and organizational need and the growing #data and technology economy presents ongoing domestic and global challenges. Visit our site provided in the comments or DM for details.