Artificial Intelligence (AI) in the context of academic studies and research is reshaping the way institutions handle information and knowledge. Aside from the disadvantages and unreliability of AI content (and the copyright concerns), there are also many opportunities with this technology.
On the following page you will find relevant information, updates, useful links, and a list of useful AI tools.
Last updated: February 2026
Please note: the information provided on this page is not intended to constitute a recommendation.
The use of AI tools must be done in accordance with the rules of use defined for the tool, as well as in accordance with the rules of ethics and the provisions of any law.
What is artificial intelligence?
Artificial intelligence allows computers to “do” things that require human thought: for example – learning a language and formulating texts, writing code, creating images and more. Concepts worth knowing:
- AI & API – a general name for the ability of software to perform an action that was not programmed directly. The acronym of API: Application Programming Interface – it means that access to the information of an external service is made.
- GenAI (Generative AI) – a subfield of artificial intelligence that generates new content based on patterns learned from a large set of examples.
- Agent – an AI-based system that performs actions autonomously to achieve a goal, often using GenAI to analyze information and generate content.
- Machine learning – the ability of algorithms to “learn” databases, identify patterns, and improve themselves over time.
- Prompts – the command line: for example, a request we make to a chat bot. The prompt can be refined in such a way that it will achieve the desired output.
Guidelines for using artificial intelligence tools in an academic context
The following guidelines will keep your academic writing professional, even when using AI tools.
For more information on scientific evaluation and tools for research support, contact the Reference and Information team (“contact us” form), or order guidance.
Using Artificial Intelligence poses a risk of copyright infringement (plagiarism):
Not every AI tool can indicate the references it relies on (the data sources). We will not be able to give proper credit without first knowing the original source of the data, and make sure it’s a legitimate academic source. Citing AI generated texts without credibility checks and without rephrasing will be considered plagiarism. Additionally, there are tools for detecting text generated by artificial intelligence, but their accuracy is not absolute, as the detection is probabilistic rather than certain.
How to avoid plagiarism when using AI tools?
You can use specific AI tools that were designed for conducting academic searches. In addition, it is always recommended to check that the data is indeed reliable – and that the content comes from the suggested reference (and not a different one). In addition, make sure the content comes from a verified and academically valid source. When citing the reference in the body of the work, always follow the citation rules adapted to the citation style and give proper credit. An example of a citation from ChatGPT in APA rules
Open access journals allow authors to pay after the article has been peer-reviewed and accepted. Unfortunately, this model of publication allows “predatory journal” type scams. In “journals” of this type there is no peer review, and citing from the articles published in them is an academic violation.
The most popular AI tools are based on a huge amount of data from around the web (in addition, ChatGPT in the free version has its data limited to the year 2021). “Facts” will come from forums, social media, blogs, private websites, news websites and more – without mentioning the reference source. We are also expected to receive information that is supposedly “academic” but comes from fake sources (such as predatory journals).
How to avoid academic misconduct?
In addition to the cross-checking of the information (1. Check that the mentioned reference exists, 2. Check that the content we are interested in citing appears in said reference), we must always verify the quality (the academic validity) of the reference. Read more in the libraries’ guide to scientific evaluation.
On the following table you will find recommendations for useful AI tools.
Please Note: some of the tools (or the options included in them) require subscription. There is currently no institutional subscription to any of the following tools. The use of AI tools must be done in accordance with the rules of use defined for the tool, as well as in accordance with the rules of ethics and the provisions of any law. For the page “AI Tools for Information Retrieval, Research, and Teaching” of the Medical Sciences Library, click here.
Recommended tools for literature review
Research Rabbit helps in finding articles on a chosen topic according to a relevant article uploaded by the user (or articles that the users add to a personal collection). By analyzing citations, it suggests relevant papers and creates a visual map to show relations between them.
The articles (recommended quantity: 5-10 articles) can be uploaded from a personal computer or added through the search option in Research Rabbit. It is also possible to obtain a visual representation of connections between articles and researchers. To get started you will need to create a personal account,. The tool can also be synchronized with Zotero for citation management. As of a 2025 update, a significant collaboration has been added with the parallel tool Litmaps.
Main sources of information: OpenAlex, Semantic Scholar, Crossref, and other sources.
Link to the tool: https://www.researchrabbit.ai/
Litmaps helps in finding articles on specific topics by using another article (or articles) as a starting point. The tool also creates a visual map of the search results and the relations between them, similar to Research Rabbit (collaborative tools).
The main sources of information on which the tool is based are Semantic Scholar, OpenAlex, and Crossref. The default is Google Scholar, but it can be switched to Semantic Scholar.
Link to the tool: https://www.litmaps.com/
Perplexity is a chat tool that answers the users’ questions, indicating within the text the sources it uses. It is now also possible to highlight a section of text, click on “Check sources,” and see exactly where the information was taken from.
The source of the information: Information from the web and social networks, with the option to focus on academic content (via Semantic Scholar). Now includes patent search as well.
Link to the tool: https://www.perplexity.ai/
Consensus is a search engine that uses language modeling technology to extract information from articles. The search is in question form (typing a question in the search box). The search mechanism includes semantic search combined with keyword search, and also allows writing search queries using Boolean operators (AND/OR/NOT). When composing a prompt in DEEP mode, filters can be applied similarly to traditional databases.
Source of information: Semantic Scholar
Link to the tool: https://consensus.app/search/
Recommended tools for academic reading
A search engine that uses language model technology to extract information from academic papers. Searches are performed by entering a research question into the search box.
The basic version allows searching for articles and extracting specific information from them in a table format. Elicit can also generate a research report based on a research question.
Source of information: Semantic Scholar & PubMed, with weekly updates.
Link to the tool: https://elicit.com/
A tool that offers multiple functionalities, including article search, extracting information into tables, answering questions about articles, and paraphrasing. It also allows saving articles to a personal library.
Information can be extracted both from search results and from PDF files uploaded to the personal library on the site. As part of the literature review features, a Deep Review option has been added—providing more in-depth search based on an intelligent agent and a “reasoning” model.
Source of information: Over 282 million articles, primarily from Semantic Scholar and OpenAlex. There is no fixed update frequency.
Link to the tool: https://scispace.com/
App link (iPhone-compatible):
https://apps.apple.com/us/app/scispace/id6742434138
The SciSpace Chrome extension allows you to access the tool’s assistance anywhere on the web. It can be used to explain article content, summarize full articles or specific sections, or perform searches.
Scite started as a tool that classifies the sentiment of citations in academic articles as Neutral, Supporting, or Contrasting the cited work. Over time it has evolved into a platform that allows article search and includes a chatbot for writing assistance.
Main information sources: Scite is not based on a single database, but rather on articles for which the company has full-text access for data mining. Scite has signed agreements with various publishers to obtain this access.
Key features: The tool offers a free browser extension that provides information from anywhere on the web, including the number of citations of an article, their sentiment, and whether the article has been retracted. It also includes a Zotero plugin and a tool for evaluating the quality of references in a paper.
Cost: Paid (except for the browser extension).
Link to the tool: https://scite.ai/
NotebookLM is a tool by Google designed for managing information sources. It allows users to upload a variety of content, including different file types / media and links. The tool assists with summarizing and organizing content through a chat interface. A unique feature is the ability to generate a customized audio summary (podcast) based on the notebook’s content, in addition to creating presentations, media, and quizzes to support learning. The tool is available in many languages, including Hebrew.
Source of information: The user provides the source(s) on which the chat is based. Note: Do not upload copyrighted or private information. Use discretion regarding privacy and data security when uploading any content to AI tools. In addition to personal notebooks, there are also public notebooks created by Google on various topics, and it is possible to search for sources in multiple modes, including deep research.
Link to the tool: https://notebooklm.google/
Top AI tools (general)
ChatGPT (Version 5.2): The new model can understand more complex requests than previous versions and provide more relevant responses, using different models including a Deep Research model with web browsing.
This version can also “read” files, images, and links, as well as generate images and write code from prompts. There are dedicated GPTs designed for specific tasks, created by users (custom bot creation is available for paid subscribers). Account settings allow personalization such as response style, persistent prompts, parental controls, and data sharing. The technology also integrates with external tools such as Canva, Figma, and more.
Gemini (by Google): A language model based on a vast dataset compiled from across the web, providing answers on a wide range of topics. Like its competitors, it is a versatile tool that can work with images, audio, code, and text. It includes an advanced image generation and editing model (“Nano-Banana Pro”) and a video generation model. The chat can be connected to other Google services such as Gmail, YouTube, Google Drive, and NotebookLM.
Claude (by Anthropic): Rich in information and frequently updated with content from across the web. It undergoes safety testing by dedicated research teams. It includes a feature called Artifacts, which allows users to view outputs in real time. Paid subscribers have access to Projects, enabling teamwork and integration with Google Docs.
Claude Code: A dedicated tool for developers that enables direct work on codebases, writing tests, and understanding complex projects, while integrating into software development workflows.
Copilot (Microsoft): Part of a broader suite of AI-based tools and services available to Microsoft users. It offers free access to a GPT-5.1 model, as well as advanced image generation and reading capabilities. Recommended for use with the Edge browser. A paid version, Copilot Pro, is integrated into Microsoft Office applications. Registering with an institutional email (e.g., from the Technion) provides enterprise-level privacy.
Each of the chats mentioned here has unique advantages and disadvantages.
It is recommended to experiment with the tools and choose according to your needs.
Image generators are AI tools that allow users to easily create images by prompting a request.
The most popular tools: Nano Banana by Gemini, Sora, Midjourney, Dall-E and Bing Image Creator
Recommended image generator: BRIA AI is an image generator developed by an Israeli company that emphasizes the issue of copyrights. BRIA AI’s algorithm is trained only on content in the public domain or content that has been approved by the original creators.
There are great AI tools that assist in proofreading. The tools can recognize the tone of voice in which the text is written (whether it is formal or direct, for example), they can recognize mistakes in grammar and spelling errors, and even rephrase long sentences, or elaborate and add text to short sentences.
Recommended tools: Grammarly, Trinka.ai and Wordtune
Grammar Checker in ChatGPT
A dedicated GPT for correcting writing errors. You can upload text and request corrections. The text will be revised with highlighted edits, along with explanations. Works in multiple languages.
Recommended tools for the Faculty of Medicine
An NLM tool that allows you to formulate the search query in the form of a sentence and search for similar sentences in articles. Finds sentences with at least 60% similarity.
Source of information: PubMed and full-text articles from PubMed Central.
Links to the tool:
Developed in the Netherlands. Searches for data in a chat platform or by PICO fields. Searching through the chat provides summarized information with references as an answer to the question, with the option of a follow-up question.
The information can be displayed in various formats, such as a table or a post tailored to a specific social media platform. Search results can be filtered by Impact Factor or sample size. When using PICO-based search, articles can be refined according to cost-effectiveness.
Source of information: PubMed, with daily updates
Link to the tool: https://evidencehunt.com/
Video tutorial: recording link
OpenEvidence is a text analysis tool with an emphasis on medical information. When you enter your research question in the search field, the tool offers its edits as needed, and produces a summary of the answer with references. You can also ask follow-up questions according to the tool’s suggestion or by choice. There is also an option for deep search (“Deep Consult”) and for uploading files.
Source of information: articles from high-ranking journals indexed in PubMed, and American clinical guidelines. They have recently signed content supply agreements with NEJM as well as JAMA.
Link to the tool: https://www.openevidence.com/
App: Available for iOS devices and Android devices.
Note: Use is restricted for the general public but free for healthcare professionals. Free access can be obtained by verifying with an Israeli license number. If verification is not approved in this way, you can contact: help@openevidence.com to arrange access.
OpenEvidence הוא כלי לניתוח טקסט עם דגש על מידע רפואי. בתיבת החיפוש כותבים את שאלת המחקר הרצויה, הכלי מציע עריכות שלה לפי הצורך, ומייצר סיכום של התשובה עם רפרנסים. ניתן לשאול גם שאלות המשך על פי הצעת הכלי או לפי בחירה. קיימת גם אפשרות לחיפוש עמוק (“Deep consult”) ולהעלאת קבצים.
מקור המידע: מאמרים מכתבי עת בעלי דירוג גבוה המאונדקסים ב – PubMed, והנחיות קליניות אמריקאיות. לאחרונה חתמו על הסכם אספקת תוכן עם NEJM וכן עם JAMA.
קישור לכלי: https://www.openevidence.com/
אפליקציה: אפליקציה למכשירי IOS + אפליקציה למכשירי Android.
שימו לב: השימוש מוגבל לציבור הכללי אך חופשי לאנשי מקצועות הבריאות. ניתן לקבל גישה חופשית בהזדהות עם מספר רשיון ישראלי. במידה והזיהוי לא מאושר בצורה זו ניתן לפנות לדוא”ל help@openevidence.com להסדרת הגישה.
Legal issues and further reading
Legal issues (in Hebrew):
- Principles of policy, regulation and ethics in the field of artificial intelligence, Ministry of Justice of Israel (Last update: December, 2023)
- The Center for AI Regulation and Policy – Ministry of Innovation, Science and Technology
- Italy Enacts the First National Artificial Intelligence Law in the European Union – Complementing the AI-Act
- Artificial Intelligence in Israel: Legislative Proposals
- The District Court ruled: Artificial intelligence cannot be considered an “inventor” under patent law
Ethics in Artificial Intelligence – Important:
Discriminations and biases based on gender, skin color, age, geographical location, and more are common because the system is not able to make decisions based on human parameters such as morality (the “black box” problem of AI). In a world where AI has a major impact, humans must apply critical thinking, whether they are AI experts or decision-makers in other fields.
Additional sources (in Hebrew):
- Chen, Z. Ethics and discrimination in artificial intelligence-enabled recruitment practices. Humanit Soc Sci Commun 10, 567 (2023)
- Osher, M., & Barak, M. (2024). Promoting ethics in the development and use of artificial intelligence among engineers and scientists of the future TheMarker Labels, trends in education.
- Kabir, A. (2021). On the relationship between artificial intelligence, human intelligence, ethics and morality Calcalist-Tech, a technological report
- Schwartz Altshuler, T., & Kahana, A. (2023). Man, Machine, State: Towards an Order of Artificial Intelligence The Center for Democratic Values and Institutions, The Israel Institute for Democracy
- UNESCO Recommendation on the Ethics of Artificial Intelligence
Your Recommendations
Have you come across professional updates relevant to this page? Are there any excellent AI tools you’d like to recommend?
We would be happy to receive recommendations for content sources, tools, and AI-based extensions, whether field-specific or general. Feel free to contact us
2026 / Reference & Information department at The Elyachar Central Library, in collaboration with the Alfred Goldschmidt Medical Sciences Library