Checkpoint 2.5

2.5 — Assessment & Grading in an AI EraFrontier

What this is

Rethinking evaluation — rubrics, authentic tasks, process-based assessment, standards-based grading — so that student learning (not AI output) is what is measured. Distinct from academic integrity (2.4): this is about what counts as evidence of learning.

Why it matters

Most assignments were designed assuming students worked alone with limited resources. AI changes that assumption. Assessment redesign keeps grades measuring learning rather than AI proficiency and protects fairness for students with and without easy access to AI tools.

Connects to

The Framework: Cognitive & Ethical Foundation — Knowledge Building & Retention; Condition #11 (Cognitive Counterweights).

Maturity levels

Not Started
Assessment practices have not been reconsidered in light of AI. Grades remain based on final product alone. AI use is either ignored in assessment design or treated only as an integrity issue. This reflects the current state of the field — AI-era assessment design is still being developed — more than a district-specific gap.
Emerging
Some teachers experimenting with process-based assessment. No district approach. Some subjects (ELA, history) ahead of others (math, science) in adaptation.
Established
District-level conversation about authentic assessment. PD on process portfolios, oral components, and performance tasks. Rubrics emphasize thinking and learning, not just product polish. Grade-band differentiated. Integrated with academic integrity (2.4).
Expanding
Assessment redesigned for AI era: evidence of learning prioritized over evidence of output. Multiple modes of demonstration (written, oral, visual, demonstrated). Students reflect on their own thinking process as part of assessment. Aligned with state assessment shifts and competency-based frameworks where applicable.

Go deeper with

Example resource
Understanding by Design (Wiggins & McTighe)
Also consider