The recent blizzard of warnings about artificial intelligence and how it is transforming learning, upending legal, financial and organizational functions, and reshaping social and cultural interaction, have mostly left out the role it is already playing in governance. Governments in the US at every level are attempting the transition from a programmatic model of service delivery to a citizen-focused model. Los Angeles, the US's second largest city, is a pioneer in the field, unveiling technologies to help streamline bureaucratic functions from police recruitment to paying parking tickets to filling potholes or locating resources at the library. For now, AI advances are limited to automation. When ChatGPT was asked recently about how it might change how people deal with government, it responded that "the next generation of AI, which includes ChatGPT, has the potential to revolutionize the way governments interact with their citizens."

An AI art lecturer said he believes the U.S. government would encounter difficulty if it attempted to establish a watermark system for AI-generated content. The U.S. will likely have a tough time trying to regulate AI-generated content, such as requiring watermarks on computer-made media, a university art lecturer told Fox News. "[F]or us to enforce it would be a lot more difficult," Tyler Coleman, who teaches University of Texas classes focused on AI, said. "I think it will be harder to achieve in the U.S. than it would be in China." China's government announced regulations in December 2022 requiring any AI-generated content to include a flag such as a watermark to indicate its origin.

The rise of artificial intelligence is entrenching more economic and political power in the hands of Big Tech companies, according to researchers at New York University (NYU) who argue AI must undergo more scrutiny and regulation. AI Now, a research institute at NYU, released a new report detailing how major tech companies wield significant control over AI, arguing such influence must be addressed now before the situation gets too out of hand. "This report is written with this task in mind: we are drawing from our experiences inside and outside government to outline an agenda for how we -- as a group of individuals, communities, and institutions deeply concerned about the impact of AI unfolding around us -- can meaningfully confront the core problem that AI presents, and one of the most difficult challenges of our time: the concentration of economic and political power in the hands of the tech industry -- Big Tech in particular," the document states. The authors add that AI development has been "foundationally reliant" on resources controlled by Big Tech, including data and computer power. Plus, they write, Big Tech companies have gained geopolitical importance by playing a central role in the U.S.-China race for AI supremacy, thereby conflating "the continued dominance of Big Tech as synonymous with U.S. economic prowess, and [ensuring] the continued accrual of resources and political capital to these companies."

The release of ChatGPT has uncovered a range of possibilities whereby large language models (LLMs) can substitute human intelligence. In this paper, we seek to understand whether ChatGPT has the potential to reproduce human-generated label annotations in social computing tasks. Such an achievement could significantly reduce the cost and complexity of social computing research. As such, we use ChatGPT to relabel five seminal datasets covering stance detection (2x), sentiment analysis, hate speech, and bot detection. Our results highlight that ChatGPT does have the potential to handle these data annotation tasks, although a number of challenges remain. ChatGPT obtains an average accuracy 0.609. Performance is highest for the sentiment analysis dataset, with ChatGPT correctly annotating 64.9% of tweets. Yet, we show that performance varies substantially across individual labels. We believe this work can open up new lines of analysis and act as a basis for future research into the exploitation of ChatGPT for human annotation tasks.

Prior studies in privacy policies frame the question answering (QA) task as identifying the most relevant text segment or a list of sentences from a policy document given a user query. Existing labeled datasets are heavily imbalanced (only a few relevant segments), limiting the QA performance in this domain. In this paper, we develop a data augmentation framework based on ensembling retriever models that captures the relevant text segments from unlabeled policy documents and expand the positive examples in the training set. In addition, to improve the diversity and quality of the augmented data, we leverage multiple pre-trained language models (LMs) and cascade them with noise reduction filter models. Using our augmented data on the PrivacyQA benchmark, we elevate the existing baseline by a large margin (10\% F1) and achieve a new state-of-the-art F1 score of 50\%. Our ablation studies provide further insights into the effectiveness of our approach.

Public opinion is a crucial factor in shaping political decision-making. Nowadays, social media has become an essential platform for individuals to engage in political discussions and express their political views, presenting researchers with an invaluable resource for analyzing public opinion. In this paper, we focus on the 2020 US presidential election and create a large-scale dataset from Twitter. To detect political opinions in tweets, we build a user-tweet bipartite graph based on users' posting and retweeting behaviors and convert the task into a Graph Neural Network (GNN)-based node classification problem. Then, we introduce a novel skip aggregation mechanism that makes tweet nodes aggregate information from second-order neighbors, which are also tweet nodes due to the graph's bipartite nature, effectively leveraging user behavioral information. The experimental results show that our proposed model significantly outperforms several competitive baselines. Further analyses demonstrate the significance of user behavioral information and the effectiveness of skip aggregation.

Traffic congestion caused by non-recurring incidents such as vehicle crashes and debris is a key issue for Traffic Management Centers (TMCs). Clearing incidents in a timely manner is essential for improving safety and reducing delays and emissions for the traveling public. However, TMCs and other responders face a challenge in predicting the duration of incidents (until the roadway is clear), making decisions of what resources to deploy difficult. To address this problem, this research developed an analytical framework and end-to-end machine-learning solution for predicting incident duration based on information available as soon as an incident report is received. Quality predictions of incident duration can help TMCs and other responders take a proactive approach in deploying responder services such as tow trucks, maintenance crews or activating alternative routes. The predictions use a combination of classification and regression machine learning modules. The performance of the developed solution has been evaluated based on the Mean Absolute Error (MAE), or deviation from the actual incident duration as well as Area Under the Curve (AUC) and Mean Absolute Percentage Error (MAPE). The results showed that the framework significantly improved incident duration prediction compared to methods from previous research.

Algorithms and models are increasingly deployed to inform decisions about people, inevitably affecting their lives. As a consequence, those in charge of developing these models must carefully evaluate their impact on different groups of people and favour group fairness, that is, ensure that groups determined by sensitive demographic attributes, such as race or sex, are not treated unjustly. To achieve this goal, the availability (awareness) of these demographic attributes to those evaluating the impact of these models is fundamental. Unfortunately, collecting and storing these attributes is often in conflict with industry practices and legislation on data minimisation and privacy. For this reason, it can be hard to measure the group fairness of trained models, even from within the companies developing them. In this work, we tackle the problem of measuring group fairness under unawareness of sensitive attributes, by using techniques from quantification, a supervised learning task concerned with directly providing group-level prevalence estimates (rather than individual-level class labels). We show that quantification approaches are particularly suited to tackle the fairness-under-unawareness problem, as they are robust to inevitable distribution shifts while at the same time decoupling the (desirable) objective of measuring group fairness from the (undesirable) side effect of allowing the inference of sensitive attributes of individuals. More in detail, we show that fairness under unawareness can be cast as a quantification problem and solved with proven methods from the quantification literature. We show that these methods outperform previous approaches to measure demographic parity in five experimental protocols, corresponding to important challenges that complicate the estimation of classifier fairness under unawareness.

'The CyberGuy' Kurt Knutsson gives a rundown on concerns over using artificial intelligence in the diagnosis and treatment of illnesses and scammers using the software for voice cloning. Using machine learning methods and available data when patients enter the hospital, researchers have developed a model that predicts strokes with more accuracy than current models. The team from Carnegie Mellon University, Florida International University and Santa Clara University published their work in the Journal of Medical Internet Research earlier this year. Examining more than 143,000 hospital visits of patients in Florida acute care hospitals from 2012 to 2014, and social determinants of health data – including the conditions people are born into and live in, and what drives those conditions – from the U.S. Census Bureau's American Community Survey, the study's authors sought to develop a stroke-prediction algorithm. Their model incorporated variables routinely collected by health care providers and payers, including basic demographics, the number of chronic conditions and insurance.

Homeland Security Secretary Alejandro Mayorkas on Friday announced the creation of a task force that will assess the ways artificial intelligence can be used to detect shipments of dangerous fentanyl to the U.S., screen cargo and take on other tasks aimed at shoring up U.S. national security. "I am directing the creation of our department's first Artificial Intelligence Task Force that will drive specific applications of AI to advance our critical homeland security missions," Mayorkas said Friday. "Countering the multi-faceted threat posed by the PRC, learning from major cyber incidents, and harnessing the power of AI to advance our security will draw on the entirety of the capabilities and expertise the 260,000 personnel of DHS bring to bear every single day," he said. "It will require continued investment in our operational cohesion, our ability to work together in ways our founders never imagined." HAWLEY GRILLS MAYORKAS ON REPORTS OF FORCED CHILD MIGRANT LABOR: 'WHY SHOULDN'T YOU BE IMPEACHED FOR THIS?' Homeland Security Secretary Alejandro Mayorkas said Friday his department is looking at how to harness AI to boost U.S. national security.