Faster initial retrieval of misinformation corrections predicts better long-term memory: An ERP study.

ABSTRACT The active expansion of digital media, as well as social networking services, has raised the dispersion of false information and fake news to serious issues, which put pressure on society, its opinion, and democracy itself. Manual verification is not easy and efficient since fake news tend to spread more rapidly than real news because it is sensational. The current research provides NEWSGUARD, a smart system that is aimed at the automated fake news detection relying on a hybrid method of using machine learning and deep learning techniques. The suggested system proposes the use of Natural Language Processing (NLP) in processing of text and extracting features. Term Frequency-Inverse Document Frequency (TF-IDF) is used to create statistical features whereas semantic representations are retrieved with word embeddings. Various classification models are deployed, such as Naive Bayes, Logistic Regression, Support Vector Machine (SVM), Convolutional Neural Networks (CNN) and Long short-term memory (LSTM). The hybrid model is created that combines the benefits of both machine learning and deep learning models and allows achieving better performance in classification. The experimental findings show that the hybrid model has a higher performance compared to the individual models where it has a high accuracy with a high precision, recall and F1 score. This research result demonstrates the usefulness of statistical and semantic features as a combination to detect fake news. The suggested NEWSGUARD solution offers the scalable, correct, and effective tool to fight with the misinformation, and it will be able to be further improved by using advanced models and real-time implementation.

The COVID-19 pandemic has sparked interest in identifying its repercussions on the mental health of older adults due to their susceptibility to abundant and often false information disseminated by the media, generating the phenomenon known as "infodemic." The objective of this study is to analyze the relationship between the COVID-19 infodemic and its impact on the mental health of older adults. A mixed-methods study with an explicit sequential strategy was conducted, using the Geriatric Depression Scale and the Geriatric Anxiety Inventory with 292 older adults. The study was carried out in two phases: the first involved data collection from older adults using the Forms platform, followed by inferential statistics analysis using Excel and SPSS to correlate variables; the second phase consisted of a qualitative study. The results showed that greater exposure to COVID-19 information from the media was associated with depression and anxiety in older adults, with a p-value of .000. Significant data, origin, gender, studies. It can be concluded that health crises affect the mental health of older adults, impacting various resources. It is crucial to understand these repercussions to implement mitigation strategies, improving the quality and accuracy of information through public policies and communication, to protect their mental health in future crises

The American Society of Clinical Oncology (ASCO) convened a multidisciplinary panel in 2017, resulting in patient-oncologist communication guidelines. Ideally, these conversations should be documented in the medical records. However, chart review for communication topics is inefficient. Large language models (LLMs) present a computational method for identification of communication domains in clinical notes, subsequently providing feedback for clinicians. The purpose of this study was to develop an approach using LLMs to identify communication domains in unstructured free text notes, validating against gold-standard chart review. The study population included 134 clinical notes from 30 patients with advanced cancer seen in June 2024 at one of seven Dana-Farber Cancer Institute clinics (Boston, MA). We used a HIPAA-secure artificial intelligence tool based on GPT-4o to develop an LLM prompt for identification of communication domains. We used standard performance metrics to compare the LLM prompt to chart review for all six communication domains. A hallucination index was calculated to assess false information that may be produced by LLMs when applied to large data sets. Across communication domains, compared to chart review, the note-level LLM analysis achieved sensitivity ranging from 0.43 to 1.0, specificity ranging from 0.32 to 0.99, and accuracy ranging from 0.51 to 0.99. The average hallucination index for all domains was low. LLM abstraction required approximately 7 seconds per note, compared to 5-7 minutes with chart review. LLMs have the potential to identify ASCO communication domains. Future directions include applying the method for quality improvement efforts, such as generating feedback for oncologists on topics that may require follow-up.

The capabilities of deepfake technology are boosted by artificial intelligence and deep learning to transform digital media. This technology generates artificial content presentation that often resembles genuine substance. Deepfake technology enables benefits in accessibility along with education and entertainment systems but dangerous utilization creates significant concerns. Artificial intelligence and deep learning enable three major risks: false information dissemination combined with identity theft and cybercrime activities. The research analyzes deepfake technology and demonstrates its operational mechanisms while evaluating its advantages and disadvantages. The research evaluates security dangers together with moral concerns while supplying real-world illustrations that demonstrate its societal influence. Various detection approaches including blockchain verification and AI-based models together with legal instruments for stopping deepfake crimes are examined in the paper. We must maintain creative development and ethical practices for deepfake technology evolution to minimize possible risks while unlocking maximum positive usage opportunities. Keywords: Deepfake, Generative Adversial Networks, Artificial Intelligence, Autoencoders

Abstract — The rapid growth of social media and digital platforms has made it extremely easy for false information, commonly known as fake news, to spread at a massive scale. This misinformation causes serious damage to society, affecting political decisions, public health responses, and general trust in news sources. Traditional methods of manually fact-checking news are slow and cannot keep up with the huge volume of content being published every day. This paper presents a Fake News Detection System that uses Machine Learning (ML) and Natural Language Processing (NLP) techniques to automatically identify whether a given news article is real or fake. The system uses TF-IDF (Term Frequency-Inverse Document Frequency) for converting text into numerical features, and applies two classification algorithms — Logistic Regression and Naive Bayes — to predict the authenticity of news. The model is trained and tested on publicly available datasets. Experimental results show that Logistic Regression achieved an accuracy of approximately 93%, while Naive Bayes achieved around 89%. The proposed system provides a fast, scalable, and reasonably accurate approach to tackling the growing problem of fake news in the digital age. Keywords — Fake News Detection, Machine Learning, Natural Language Processing, TF-IDF, Logistic Regression, Naive Bayes.

ChatGPT is a popular artificial intelligence (AI) tool used to answer questions on any subject. Given ChatGPT's popularity, it is prudent to investigate its ability to answer common patient questions in the field of hand therapy to better guide patients as they navigate the resources available to them. This is a cross-sectional, rater-based comparison study. Four common hand therapy questions were entered into ChatGPT version 3.5. The first five answer tabs that appeared with a Google search for the same four questions were downloaded. Three certified hand therapists blindly graded ChatGPT and Google's answers using Likert scales to assess for answer accuracy (0-6), comprehensiveness (0-3), and conciseness (0-3). ChatGPT was significantly more accurate, with an estimated marginal mean (EMM) of 5.75 (95% CI: 4.96, 6.54) compared to Google's 3.48 (95% CI: 2.86, 4.10) (p < 0.001). ChatGPT was significantly more complete, with an EMM of 2.50 (95% CI: 2.10, 2.90) compared to Google's 1.48 (95% CI: 1.19, 1.77) (p < 0.001). ChatGPT was significantly more concise, with an EMM of 3.00 (95% CI: 2.66, 3.34) versus 1.60 (95% CI: 1.29, 1.91) for Google (p < 0.001). ChatGPT is a concise, comprehensive, and accurate alternative to a Google search for people seeking information on hand therapy. The free version of ChatGPT does not update its sourcing past 2019, and the software is known to occasionally present false information. Frequently updated academic websites should therefore remain the primary online medical resource for patients.

Facing the problem that social robots cause false information to spread, this paper constructs an information diffusion prediction model for dynamic behavior recognition. The model combines deep learning with spatio-temporal modeling technology, aiming at improving the recognition accuracy of social robots with strong camouflage and realizing the dynamic prediction of their information propagation paths. In this study, a multimodal recognition model based on a graph attention network (GAT) and a bidirectional long-term and short-term memory network (BiLSTM) is built, and the time sequence characteristics of user behavior, the semantic content of text and social network structure are modeled by combining them. Based on this, a spatio-temporal graph convolution network (ST-GCN) is designed to model the information diffusion process, capture the spatio-temporal dependence in communication, and predict the future diffusion trend. The results show that the F1-score of the proposed recognition model reaches 0.892, which is about 8% higher than that of the traditional graph convolution network (GCN). In the diffusion prediction task, ST-GCN can respond to sudden propagation events more accurately. This study provides a technical route with dynamic perception and prediction ability for social platforms.

Abstract Humans are endowed with a suite of cognitive mechanisms that enable them to mitigate the risk of misinformation and underlie their epistemic vigilance. When direct access to the initial source of information is unavailable, individuals often rely on the vigilance of others to acquire beliefs. Moreover, monitoring the critical alertness of one’s interlocutors is essential for interpreting their communicative intentions, particularly in cases of deliberately conveying false information, such as lies or verbal irony. However, little is known about the human capacity to track others’ epistemic vigilance and use this information to guide belief formation and pragmatic interpretation. This study investigates whether children aged from 4 to 8 selectively trust vigilant over gullible informants. Our findings suggest that selective trust based on informant vigilance towards inaccuracy begins to emerge around the age of 6.

The main task of a Notary is to make authentic deeds, of course this has consequences where the Notary who makes the authentic deed has responsibility for the authentic deed he makes in order to provide legal certainty to the public in order to provide legal protection. This research uses a doctrinal approach method. The first result of this research is, the notary's responsibility for the deed he has made, if it is proven to contain false information and the notary knows about it, the notary must be responsible both from an administrative, civil law and criminal law perspective for the deed he has made. If the notary does not know that the information is false information, then the notary concerned is not responsible for the false information contained in the deed he made. A third party who suffers loss as a result of an authentic deed that is suspected of containing false information made by the notary must prove that there are untruths in the contents of the authentic deed so that he or she can receive legal protection. If the third party does not have evidence that he suffered losses from an authentic deed which is suspected of containing false information made by a notary, then he does not receive legal protection. Because an authentic notarial deed has perfect evidentiary power as regulated in Article 1870 of the Civil Code. Second, Notaries receive preventive and repressive legal protection through the establishment and enforcement of laws. The Notary Honorary Council (MKN) plays a role in providing approval for the summons and retrieval of notary documents in the judicial process, as well as providing guidance to maintain the integrity of notaries.

False information can hinder individuals’ cognitive processes for news verification, potentially confusing members of the public and sowing distrust in information sources. The maxim “trust, but verify” has long been commonly used in information processing, but the updated maxim “verify, then trust” proposes a principle by which people first evaluate the truthfulness of the information they come across by examining its source and seeking validation from authoritative experts. Based on a survey of U.S. adults, the author examines the role of the need for cognition (NFC) on perceived exposure to false information (PEFI) and news verification behaviors, in combination with the strategy, “verify, then trust.” The results of a serial mediation analysis suggest that, while PEFI is linked to decreased trust in media, NFC can have the opposite association when mediated by PEFI and news verification, offering broader insight into the significance of cognitive thinking in news consumption behaviors. By integrating multiple theoretical frameworks, this study highlights the joint endeavors of audiences and journalists.

This research explored how artificial intelligence health content creation affects patient trust toward South-West Nigerian medical establishments. The fast growth of AI-based tools creates new business opportunities, but also brings operational difficulties, which affect how doctors interact with their patients and how patients trust their healthcare providers. The research used a mixed-methods approach that began with purposive qualitative interviews with 25 patients who received care at general hospitals in Lagos, Oyo, and Ekiti States to study their experiences with AI-generated medical information. The research showed that trust develops through personal human connections. Yet people see AI content as both helpful for access and dangerous because it creates false information. The quantitative phase involved a survey administered to 150 participants to evaluate the themes and hypotheses that emerged from the initial qualitative findings. A multiple linear regression analysis was conducted, finding that the model was statistically significant (p&lt; .001) and explained 65% of the variance in patient trust. The analysis identified three key variables significantly influencing trust: perceived AI accuracy (beta = 0.32), perceived risk of misinformation (a strong negative predictor, beta = -0.51), and digital literacy (beta = 0.40). The research concludes that a significant relationship exists, confirming that while AI offers a new channel for information, trust in healthcare remains fundamentally tied to human interaction and the ability of patients to critically evaluate digital information. Based on these insights, recommendations are provided to help institutions build digital literacy, manage misinformation risk, and strategically integrate AI to reinforce patient confidence.

In modern conditions, information flows are complicated by unreliable and harmful data, which negatively affects management systems and information security. Tools are needed to filter information before it is processed. The increase in false and malicious messages requires effective algorithms for analyzing and managing data that ensure the stability of automated systems. The purpose of the research is to create effective mathematical and computational methods for the analysis, classification and management of information to improve the reliability of systems and the reliability of data. A method of simulation modeling based on a mathematical model with elements of probability theory is proposed, where the information flow is divided into reliable, false and harmful information. To classify messages, probabilistic methods are used, taking into account prior and posteriori probabilities, as well as the analysis of network, temporal and semantic characteristics. Unlike existing methods, this one focuses on analyzing data before it is used, which reduces the risk of destructive impacts. A mathematical model has been developed for the analysis of information flows, including reliable, false and malicious information. The model uses probabilistic approaches and considers the network, temporal and semantic characteristics of messages to classify them and minimize their destructive impact. The model allows you to effectively consider the characteristics of each source, distinguishing reliable, false and malicious messages, which ensures high accuracy and reliability of the resulting information flow. This end-to-end solution improves data integrity and can be used in management and information security systems to minimize the impact of disruptive information and enable informed decision-making. The results can be used to monitor information threats, filter malicious information and ensure the security of critical systems, as well as support decision-making in government agencies, the economy and energy, increasing trust in information systems.

Purpose This study explores the potential of Natural Language Processing (NLP) tools, specifically ChatGPT-4o and Data Analyst, in supporting health commissions with disability assessments. Methods Nine realistic patient scenarios were created to reflect typical cases in disability evaluations, encompassing conditions like stroke, cerebral palsy, traumatic injuries, and hereditary disorders, each involving varied motor and functional impairments. These scenarios were input into ChatGPT-4o and Data Analyst. Their outputs were evaluated using a 5-point Likert scale on alignment with expert guidelines, completeness, and lack of false information. Interrater reliability was established before scoring. Results Both models produced generally high-quality narrative reports, ChatGPT-4o was rated “very good” and Data Analyst “good,” with no statistically significant difference in overall scores. However, ChatGPT-4o failed to calculate correct disability percentages in 55.6%, and Data Analyst failed in 88.9% of scenarios. Conclusions While NLP tools can assist in generating structured disability reports, they currently lack the precision needed for calculating disability percentages reliably. Expert oversight remains essential for decision-making in disability assessments.

With the development of the global net, politicians are increasingly using the Internet to create texts with deliberately false information to form a distorted picture of the world in the mass consciousness. As a result, it leads to communicative conflicts. The aspects of implementation of conflict-generating communication in English political propaganda media discourse are analysed in this article. The purpose of the study is to determine the communicative and pragmatic qualities of linguistic units typical for the production conflict narratives within English linguistic culture in the virtual space. The posts of EU High Representative for Foreign Affairs and Security Policy Kaja Kallas and the comments of the users were taken on X. The main research methods were descriptive and contextual analyses, as well as discourse analysis. In the course of the work it was determined that the addresser uses certain manipulative tactics to introduce specific ideas and views in order to form the necessary behaviour of the addressee. The “own” — “alien” opposition is a pragmatic dominant in strengthening socio-political confrontation and achieving a perlocutionary effect. User comments are a reaction to the initial post – they reflect the dynamics of the audience’s mood and contribute to the expansion of the field of confrontation. The analysis of the material made it possible to distinguish the concept of “hate speech” as the leading method to create the narrative of the conflict within English political propaganda media discourse. This phenomenon verbalises a certain type of discrimination.

This article investigates the deception-eliminating design (DED) against false information attacks by deceptive agents in asynchronous optimal consensus control. We model the asynchronous interactions among agents as multistage games and establish a Tit-for-Tat rule to compel deceptive agents to turn to transmitting true state information. Furthermore, we design a false information counterattack rule under asynchronous updates by leveraging the invariance of the rank of equivalent matrices and the convexity of positive semidefinite quadratic forms. This design effectively intimidates deceptive agents that have transitioned to cooperation, ensuring they do not revert to transmitting false information. Subsequently, by utilizing the properties of Riccati differential equations, the integrating factor methods, the Minkowski inequality, and proof by contradiction, we theoretically analyze the impact of false information on consensus and provide an explicit upper bound for the strategy update periods of agents with different performance matrices. Theoretical proof shows that as long as the strategy update periods of all agents remain below this upper bound and the above two DEDs are implemented, the asynchronous consensus is guaranteed.

Can large language models provide accurate and empathetic answers to the most frequently asked questions by infertile patients? A pilot study.

Artificial intelligence (AI) chatbots are increasingly used in healthcare, but their ability to interpret anesthetic monitoring data during intraoperative crises remains unclear. To evaluate AI chatbots' responses to simulated anesthetic emergencies, with a focus on visual monitor interpretation compared to contextual case information. This simulation-based study was conducted using a high-fidelity patient monitor. Twenty intraoperative emergencies were designed as static monitor images with minimal clinical context. Five AI platforms, ChatGPT, Claude, Gemini, Copilot, and DeepSeek, were tested. Each scenario was submitted in a separate, new conversation with no advanced reasoning tools enabled. Six blinded evaluators scored 100 responses using the validated CLEAR tool. The primary outcome was the mean total score per chatbot; secondary outcomes included domain-specific scores and differences between visual and contextual scenarios. ChatGPT achieved the highest mean score (4.4 ± 0.6), outperforming other platforms across all CLEAR domains (P < 0.001). Its accuracy was consistent between contextual (4.6 ± 0.3) and visual (4.1 ± 0.9) scenarios. DeepSeek scored the lowest overall score (2.7 ± 1.1), and had a mean 2.3 ± 1.1 in "lack of false information", often due to misinterpreting monitor values, reducing its visual scenario performance. Overall, contextual scenarios scores were higher than visual ones across all platforms. Among AI chatbots, ChatGPT demonstrated the most consistent and guideline-concordant responses to simulated anesthetic crises as of March 2025. This study provides a benchmark for evaluating clinical AI performance and supports the selective integration of such tools in anesthesia decision-making workflows.

The exponential growth of online social networking, the proliferation of fraudulent and bot-driven accounts has emerged as a critical threat to platform integrity. These accounts are commonly exploited to disseminate false information, manipulate user behavior, and engage in various deceptive practices. Addressing this challenge requires a robust and intelligent detection mechanism capable of adapting to increasingly sophisticated evasion tactics. This paper introduces the Social Media Fake Account Detection and Prevention System (SMFADPS), a multi-layered analytical framework that assesses account authenticity through the evaluation of multiple behavioral signals. The system examines profile completeness, content credibility, follower growth irregularities, and repeated content patterns to generate intermediate suspicion scores across four specialized detection modules. An ensemble-based weighted scoring mechanism consolidates these scores into a unified risk rating, which is subsequently used to categorize accounts into distinct threat levels. The system is developed using Python, FastAPI for RESTful service delivery, and a dual-database configuration comprising PostgreSQL and MongoDB. Evaluation conducted on a large-scale dataset confirms that the multi-signal approach yields substantially higher detection accuracy and operational efficiency compared to conventional single-indicator systems.

The quick development of digital communication platforms has opened up new areas where people’s behavior is profoundly and frequently shaped by online anonymity. This paper looks at the psychological understanding of online anonymity and how it can lead to harmful behaviors like hate speech, harassment, cyberbullying, and flaming. In order to provide a thorough understanding of why people behave differently when protected by anonymity in digital environments, the study shows findings from theoretical literature, drawing on important theoretical frameworks. The research paper shows three main factors of toxic online disinhibition: anonymity, invisibility and lack of eye-contact. The long-held belief that toxic behavior is solely motivated by anonymity is challenged by experimental findings showing that the single biggest cause of negative disinhibition is absence of eye-contact. The study also shows how cyberbullying appears in online forums, how anonymity is adversely correlated with aggressive attacks, and how algorithmic amplification and social media platform design amplify these effects. The study recognizes the dual nature of anonymity in addition to the risks: although it encourages toxic behavior, it also decreases barriers to self-disclosure, assists vulnerable people seeking mental health support, and creates positive anonymous networks. The effects of toxic online behavior are examined at the individual level, such as despair, anxiety, and suicidal thoughts among victims, as well as at the communal and societal levels, such as the dissemination of false information and heightened social division. AI-powered content filtering, digital literacy instruction, identity verification systems, and platform design reform are some of the methods to lessen toxic behavior that are covered.