Everywhere you turn, technology companies are incorporating artificial intelligence (AI) into their products and services. AI has been around for decades, but the recent proliferation of low-cost, widely-available generative AI and large language models has led to accelerated adoption of AI in many sectors. We are excited to share more information about OpenScholar’s approach to AI.

OpenScholar's principles of AI

Using AI raises ethical, legal, and practical questions. The OpenScholar team has been discussing how we will incorporate AI-assisted features into our platform to best serve our customers and help them achieve the mission of turning research into impact. We’ve established six principles that will guide the work of all our teams, from sales and customer success to product and engineering.

These principles were inspired by the eight principles Pendo has shared on their blog. Since we are also in the business of developing software, many of their principles apply to us. We have modified and extended their examples to develop a set of principles unique to OpenScholar.

1. Customer-centric approach

Our goal is to provide a superior experience to the users who manage and visit OpenScholar websites. Any generative AI-powered features will be designed to improve the usability of our systems and enrich the experience for customers and for the audiences who visit their sites.

2. Transparency and communication

We will openly communicate about the generative AI-enhanced features that we develop in the OpenScholar platform. The interface will indicate when user-entered data will be shared with AI tools. We recognize that our academic users are especially sensitive to issues around plagiarism and intellectual property and our healthcare users must protect information about patient health and research subjects.

3. Optionality and customization

We realize that users have different preferences when it comes to interacting with generative AI systems. Whenever possible, we will build in options to choose whether or not to share data with AI. Organization leaders may enable some features at the institution level, and some options may be available at the individual user level.

4. Privacy and security

We will be responsible stewards of data that we collect, store, and use in our AI systems. We will respect user privacy and comply with any applicable laws, regulations, or standards surrounding the use of AI.

5. Equity and inclusion

Generative AI models inherently reflect the biases of the humans who create them. We will make efforts to identify and mitigate these biases in datasets and models that we develop so that the experience of our platform is equitable and inclusive. When integrating with third-party models, we will seek systems that also work to mitigate biases.

6. Active engagement

AI technologies are rapidly evolving. Our team is committed to staying actively engaged in learning about innovations and regulations in this area so that we can continue to bring the best experiences to our customers and their communities.

OpenScholar's AI roadmap

We envision incorporating generative AI into our platform and services in five main ways.

1. Writing assistance

Adding tools that will help content creators generate text, adjust tone and length, and add metadata to help with searching and cross-references.

2. Image assistance

Adding tools that will help content creators generate or find images and generate alternative text for accessibility.

3. Site creation assistance

Automatically populating new sites with content based on information provided during the site creation process.

4. Platform assistance

Suggesting actions or help documents in context, based on a user’s current actions, improving the onboarding process, and adding a chatbot to help with questions.

5. AI-powered search

Using natural language questions to find connections between people and labs in the OpenScholar network and surface opportunities for funding and collaboration.

Check our News & Updates section or follow us on social media for announcements about the rollout of these features.

Follow OpenScholar