Ought

San Francisco, California, US All United States companies → — Founded 2017

AI Research Lab

Family Medicine Internal Medicine Oncology

Clinical Decision Support & Reference Drug Discovery & Research

Seed (for Elicit spin-off) — $9M (for Elicit spin-off)

Website Claim this company

Products

Elicit

Ought

Clinical Decision Support & Reference

Elicit is an AI research assistant that helps physicians and researchers streamline literature reviews, extract key information from papers, and synthesize findings for research and clinical decision support.

Freemium

See all 1 product →

About Ought

Ought is a product-driven research lab focused on developing advanced machine learning systems to scale up high-quality reasoning. Their core mission is to enable AI to assist with open-ended thinking and complex decision-making, rather than solely tasks with clear, short-term outcomes. Ought spun off Elicit, an AI research assistant, which is now an independent public benefit corporation.

Elicit, incubated at Ought, is an AI-powered research assistant specifically designed to automate and streamline scientific research workflows. For physicians and researchers, Elicit offers capabilities such as finding relevant academic papers, summarizing findings, extracting key information, and organizing data into concepts. It aims to significantly reduce the time spent on tasks like systematic literature reviews, paper screening, and data extraction, with some users reporting up to an 80% time saving. Elicit emphasizes transparency, providing sentence-level citations for all AI-generated claims to ensure verifiability. The platform is built on an architecture that supervises reasoning processes rather than just outcomes, which Ought believes is crucial for supporting open-ended reasoning and ensuring AI alignment. Elicit is used by researchers at institutions like Harvard, MIT, and Stanford, as well as companies such as Genentech and Novartis.

While Ought’s broader mission encompasses various applications of advanced machine learning, Elicit’s focus on scientific research directly impacts the healthcare sector by accelerating the synthesis of medical literature and evidence-based reasoning. This can aid physicians in staying current with the latest research, conducting rapid reviews for clinical decision-making, and potentially identifying new insights from vast amounts of published data more efficiently.

Focus Areas

AI research machine learning natural language processing AI safety research automation literature review systematic review evidence synthesis data extraction

Business Intelligence

Partnerships	MicroGenDX (uses Elicit across Biostatistics, Medical Affairs, and Marketing teams)
Acquisitions	Ought has spun off Elicit as an independent public benefit corporation
Technology	Semantic search over academic databases, enhanced by LLM-powered summarization and extraction; uses machine learning models like GPT

What Physicians Need to Know

Specialization

Ought, through its product Elicit, specializes in AI-powered research assistance, primarily focusing on automating and streamlining literature reviews and evidence synthesis for academic and scientific research.

Key Differentiators

Elicit's key differentiators include its ability to perform semantic searches over a vast database of academic papers (over 138 million papers and 545,000 clinical trials), extract specific data points into customizable tables, and generate concise summaries of research. It uses a process-based AI architecture, which aims to supervise reasoning processes rather than just outcomes, and employs Retrieval Augmented Generation (RAG) to reduce hallucinations and improve factual accuracy. Elicit also offers features like automated screening and data extraction for systematic reviews, and the ability to organize and store sources.

Technology Stack

Elicit leverages machine learning models, including GPT, and works with Semantic Scholar's database of academic papers. Its architecture is based on supervising reasoning processes, and it utilizes Retrieval Augmented Generation (RAG) for improved accuracy. Ought has also developed the Interactive Composition Explorer (ICE), an open-source Python library for compositional language model programs, to demonstrate its process-based approach.

Physician Tip

For physicians, Elicit can be an invaluable tool for staying current with the latest research, quickly synthesizing evidence for clinical decision-making, and supporting evidence-based practice. It can significantly reduce the time spent on literature reviews, allowing for rapid access to key findings and data extraction from numerous studies. Physicians can use it to answer specific research questions, understand the scope of existing literature, and identify foundational articles in their areas of interest. However, it's crucial to remember that Elicit is an assistant, not a replacement for critical appraisal; always verify AI-generated information with the original sources, especially given the potential for bias and inaccuracy in AI models. It's particularly useful for empirical research questions, such as those concerning interventions and outcomes in randomized controlled trials.

Elicit primarily integrates with Semantic Scholar for its vast database of academic papers. While it allows for exporting extracted data as CSV files for integration into other projects, specific direct integrations with other clinical or EHR systems are not explicitly highlighted in the provided information.

At a glance

Founded 2017

HQ San Francisco, California, US · United States

Team size Ought: 10 (as of August 2022); Elicit: 50+ (as of March 2026)

CEO Owen Cotton-Barratt (for Ought); Andreas Stuhlmüller (for Elicit)

Funding Seed (for Elicit spin-off) — $9M (for Elicit spin-off)

Products 1

Visit site

Products by Ought

1 product in the directory

Elicit

Ought

Clinical Decision Support & Reference

Freemium

What the Web Says

Ought, the organization behind Elicit.com, has developed an AI research assistant primarily focused on automating literature reviews and knowledge synthesis for academics and researchers. The tool aims to streamline tasks like finding relevant papers, extracting data, and summarizing findings, with a strong emphasis on accuracy and transparent sourcing. While praised for its efficiency in specific research workflows, some users note limitations in its comprehensive search capabilities across all databases and its ability to handle highly nuanced or theoretical topics.

Overall: Positive

Strengths

Automates literature reviews and data extraction, saving significant time.
Provides accurate and relevant results with sentence-level citations for verification.
User-friendly interface for easy navigation.
Helpful for getting overviews of research questions and identifying research gaps.
Strong for systematic reviews and evidence synthesis.
Continuously improving with updates to its functionality.

Limitations

May lead to inaccuracies if underlying machine learning models are flawed.
Can result in information overload due to the volume of data processed.
Limited contextual understanding for nuanced or complex topics.
Primarily searches the Semantic Scholar database, potentially missing papers from other major databases.
Limited customization options.
Some users report that results can be obscure or from lesser-known journals.

Based on reviews from: Tenere (Elicit AI Reviews), Academic Help, Reddit, Skywork, Manusights, G2, AI Flow Review, Ought, Capterra, medRxiv, PMC, The National Law Review, Indeed.com

Last updated: 2026-09-11

Videos

News, demos, and interviews about Ought.

Loading videos...

Videos load interactively. If none appear, search YouTube for Ought →

View all on YouTube

Press & Coverage

Ought

Ought has spun off Elicit

Ought, a non-profit AI research organization, spun off its core product, Elicit, as an independent public benefit corporation in September 2023. Most of Ought's staff transitioned to Elicit, which subsequently raised $9 million in seed funding.

2023-09

Elicit

Elicit Raises $9 Million and Becomes a Public Benefit Corporation

Elicit, an AI research assistant developed by Ought, announced in September 2023 that it raised $9 million in seed funding and became an independent public benefit corporation. The funding was led by Fifty Years and included participation from various angel investors with expertise in AI and startups.

2023-09

Therapeutics Initiative, UBC

Elicit: Applying language models to evidence synthesis

In July 2022, Jungwon Byun, COO and cofounder of Ought, presented on Elicit, an AI research assistant that uses language models to automate parts of literature review and evidence synthesis. The presentation highlighted Elicit's ability to find relevant papers, extract key information, and summarize findings.

2022-07

IntuitionLabs

Elicit AI Data Extraction Guide for Clinical Papers - IntuitionLabs

This guide from March 2026 details how Elicit, developed by Ought, functions as an AI research assistant for structured data extraction from clinical research papers, particularly for systematic reviews. It emphasizes Elicit's use of large language models and semantic search to streamline research workflows.

2026-03

PMC (NIH)

Comparison of Elicit AI and Traditional Literature Searching in Evidence Syntheses Using Four Case Studies - PMC

A peer-reviewed article from 2025 compared Elicit AI to traditional literature searching methods in evidence syntheses, finding that while Elicit identified unique studies and offered high precision, its sensitivity was lower, suggesting it's a useful adjunct rather than a replacement for traditional methods.

2025-09

LinkedIn (David Thaw, PhD, JD/MBA)

Elicit and the End of Manual Evidence Matrices in Regulatory...

A May 2026 article discusses Elicit's potential to revolutionize regulatory compliance by automating evidence matrix creation from scientific publications and trial reports, highlighting its role in healthcare teams.

2026-05

Stork.AI

Elicit Review (2026): Pricing & Alternatives | Stork.AI

This 2026 review of Elicit highlights its features as an AI-powered research assistant for academic and pharmaceutical industries, including intelligent literature search across over 138 million papers and 545,000 clinical trials, and adherence to PRISMA 2020 guidelines.

2026-09

PMC (NIH)

Ten simple rules for optimal and careful use of generative AI in science - PMC

An October 2025 article discusses the responsible use of generative AI in scientific research, citing Ought's Elicit as an example of a tool that employs generative AI to support literature reviews by automating the extraction of key findings and summaries.

2025-10

Frequently Asked Questions

What specific AI products does Ought offer for healthcare, and how do they benefit physicians and administrators?

Ought is a product-driven research lab focused on 'scaling up good reasoning' with advanced machine learning systems. While Ought itself has spun off Elicit, which is an AI research assistant, their core mission is to automate and scale open-ended reasoning. For healthcare, this translates to AI that can support complex decision-making, potentially aiding in tasks like broad and deep literature reviews for research, and other general-purpose reasoning workflows. The goal is to enhance, not replace, human judgment, by providing tools that can analyze large amounts of information, identify patterns, and offer evidence-informed recommendations, which can improve efficiency, workflow, and detection of high-risk conditions.

How does Ought ensure compliance with healthcare regulations like HIPAA and address concerns about data privacy and algorithmic bias?

Ensuring compliance and addressing concerns like data privacy and algorithmic bias are critical in healthcare AI. While specific details about Ought's compliance framework aren't readily available, general best practices for AI in healthcare emphasize establishing lawful authority for data use, clarifying data practices, and aligning them with patient expectations. This includes robust data governance, quality controls, and human oversight of AI outputs. The industry also focuses on assessing AI solutions for potential bias and disparate impact on vulnerable groups, with a commitment to transparency and accountability in how AI tools function and are developed.

What is Ought's pricing model for its healthcare AI solutions, and how does it compare to the value offered?

The pricing models for AI in healthcare are still evolving and can be complex. Generally, costs depend on project complexity, data readiness, compliance requirements, and long-term maintenance. AI solutions can potentially reduce operational costs by automating tasks and improving efficiency, but initial investments in data preparation and custom solutions can be significant. The value offered by AI is often measured in improved outcomes, reduced administrative burden, and enhanced decision-making, which can lead to cost savings in the long run.

What kind of support does Ought provide to healthcare organizations during implementation and ongoing use of their AI products?

While specific details about Ought's support structure are not explicitly outlined, successful AI implementation in healthcare requires ongoing monitoring, validation, and continuous improvement. This often involves collaboration between AI developers, healthcare professionals, and IT teams. Effective support would likely include training staff, maintaining human oversight, and continuously monitoring AI performance to ensure accuracy, safety, and effectiveness in supporting physicians and administrators.

Does Ought engage in partnerships with other healthcare technology companies or institutions?

Many AI companies in healthcare form strategic partnerships to combine technological expertise with domain-specific knowledge. These collaborations aim to ensure AI tools are grounded in clinical realities and benefit from specialized technical expertise. While specific healthcare partnerships for Ought are not detailed, their spin-off Elicit has raised venture funding, indicating a growth trajectory that often involves strategic alliances.

What is Ought's company stability and long-term vision, particularly concerning its commitment to the healthcare sector?

Ought is described as a 'product-driven research lab' focused on AI alignment and scaling open-ended reasoning. They have spun off Elicit as an independent public benefit corporation, which has secured significant venture funding. Their long-term vision is to channel the growth of AI's cognitive capabilities toward 'good reasoning' across various domains, including potentially impacting healthcare by improving the world's ability to reason about long-term outcomes and decision-making. The stability of an AI company in a rapidly evolving field often depends on its ability to innovate, secure funding, and demonstrate the practical value and safety of its products.

Related Companies

Suggest an Edit →