Anthropic Releases New AI Courses and Research
Anthropic launched three new AI fluency courses alongside research analyzing 74k educator conversations. The data provides concrete insights into how AI is being used in higher education, while the courses address some of the challenges and opportunities these findings reveal.
Key findings from the research:
- 57% of faculty chats with Claude focus on curriculum development and instructional design
- Faculty show a preference for using Claude to augment not automate their work
- While faculty rate grading as AI’s least effective application in their practice, nearly half of grading conversations were fully automated
- Faculty expressed concerns about "cognitive offload" and students becoming overly dependent on AI, emphasizing the need for students to develop foundational AI literacy skills
The newly released courses demonstrate a commitment to higher education and begin to bridge the gap between current practices and effective AI integration.
Here are some of the course highlights:
- Created three courses—for students, for teachers, and for teaching AI fluency
- Strong "human in the loop" approach that aligns with the augmentation patterns educators prefer
- 4D framework (Delegate, Describe, Discern, Diligence) offers structured decision-making for when to collaborate vs. automate
- Focus on responsible collaboration and academic integrity, particularly relevant given the grading automation concerns
- Content designed for college-level learners, though not suitable for K-12 audiences
The alignment between the research findings and course principles is encouraging. However, the data also reveals why targeted training matters—educators are innovating in curriculum development, but may need additional support around assessment practices.
It's promising to see how these frameworks align with approaches like our SEE model (Safe, Ethical, Effective AI use), pointing toward more consistent standards across the field.
Link to the courses and the research in the comments.
