If you have an item that we should include in this news download, or a source we should review for future items, please email [email protected].
EVENT – AI for Organizing & Campaigns Hackathon - On August 9th Cooperative Impact Lab will host a hackathon in New York City to develop innovative solutions for how AI can support organizing campaigns and progressive causes. Click here to read more and register.
With President Biden passing the reins to VP Harris, Fast Company (July 22, 2024) reports on Kamala Harris's stance on artificial intelligence while serving as Vice President and as California’s Attorney General. While Harris has accepted donations from prominent Silicon Valley venture capitalists, she has also pressured tech companies to support stronger privacy protections for users and to crack down on users sharing explicit images without the subject’s consent. More recently, Harris has warned of dangers posed by generative AI.
With elections in the news, the Washington Post (July 22, 2024) reports on how AI assistants have struggled to provide accurate information about recent high-profile political events – notably, the attempted assassination of former president Donald Trump and President Biden’s announcement that he was withdrawing from the 2024 presidential race. The Post asked numerous chatbots for current information about these stories as they unfolded and got varied responses; some models denied that the events were happening, while others said they didn’t have up to date information and pointed users to online news sources. This experience calls into question the role that AI assistants should play when it comes to how the public searches for and receives information about time-sensitive political news.
Wired (July 24, 2024) has a piece about the ACLU’s fight to protect the right to make deepfakes. As lawmakers across the country have introduced dozens of pieces of legislation seeking to put guardrails on the use of deepfakes – a type of AI-generated media that depicts false or misleading people or events presented as real – the American Civil Liberties Union (ACLU) is pushing back. The ACLU has opposed legislation that would allow individuals portrayed in deepfakes to sue the creators, as well as legislation requiring deepfakes portraying real people to be labeled. The legal advocacy group has argued that such regulations violate the first amendment by illegally restricting the free speech rights of the creators of deepfakes, as well as the rights of others to consume legally protected speech.
Generative AI is speeding procurement, but not without risks, says StateScoop (July 17, 2024). Some state and local governments are using generative AI to make the procurement process quicker and more efficient. From Idaho to Utah, procurement managers are using AI tools to complete large tasks such as creating requests for proposals (RFPs) more efficiently, and to better communicate with subject matter experts. At the same time, experts have highlighted risks, including bias, transparency and privacy concerns about the technology. Another consideration is that the lack of cohesion among various local, state and federal policies when it comes to generative AI could lead to compliance issues.
The local government of Shanghai issued a declaration on AI. Reported in AI Business (July 8, 2024) Shanghai committed to “promote the development and application of AI technologies while ensuring safety, reliability, controllability and fairness in the process.” The full text of the Shanghai Declaration is here. The United Nations Industrial Development Organization (UNIDO) also announced the creation of a new research center to promote the development of AI technologies for use in manufacturing industries in the Global South.
StateScoop (July 5, 2024) reports on how New Orleans 911 is triaging emergency calls with AI. The Orleans Parish Communication District, which handles the City of New Orleans’s 911 service, is experimenting with an AI software product called Call Triage to see if the tool can help staff process emergency calls more efficiently. The software helps the Parish’s digital dispatch system to identify redundant calls – in other words, incidents that dispatchers are already responding to – so that human staff can focus on fielding calls about new incidents. The district hopes that the AI system can help to relieve staffing shortages – a common problem for emergency response centers across the country.
In an interview about performance management and AI at the Social Security Administration with FedScoop (July 5, 2024), three officials – the Chief Information Officer, Commissioner, and Operating Officer – highlighted efforts to make better use of the administrative data the agency collects. The agency has implemented a software tool called SecurityStat which allows staff to measure and visualize performance metrics such as processing times for benefits and wait times for the agency’s customer service line. The interviewees say that the adoption of SecurityStat is part of a larger effort to improve how the agency uses AI and data tools to deliver services to the public.
As Government Technology (July 5, 2024) reports, the Regional Transportation Commission of Southern Nevada – the public transit agency for Las Vegas and the surrounding area – has adopted an AI-based platform created by the company ZeroEyes to aid in weapons detection. The system works by analyzing video captured by the agency’s existing digital security cameras. If the system detects a gun in the transit system, it will share the images with company staff for review who can then alert law enforcement if a threat is detected. ZeroEyes says that their tool can help to mitigate gun-related violence by reducing response times.
