Artificial intelligence applications: geared towards teachers

The fluorescent lights hummed above Elena’s desk, casting a flat glare over the stack of essays that threatened to topple onto the floor. She rubbed her temples, the pen in her hand moving slowly across the page as she wrote, for the third time that evening, “Needs more specific evidence to support the claim.” The comment felt hollow, a reflex rather than a response. Outside, the October sky had deepened to indigo, and the parking lot lights flickered on one by one.

She leaned back in her chair, the vinyl creaking under her weight. Another essay lay open—Javier Morales’s, his handwriting tight and deliberate, his argument about the Montgomery Bus Boycott thoughtful but buried under vague phrasing. She sighed. Unclear thesis, she began to write, then paused. How many times had she written that exact phrase this week? It wasn’t that Javier didn’t understand; it was that his sentences often started with prepositional phrases, a rhythm shaped by growing up speaking Spanish at home. The AI, she knew from the district memo, would flag that as a syntax issue. She didn’t need a machine to tell her that. She needed time—time to talk to him, to ask what he was trying to say.

The faculty lounge door swung open, and Marcus Chen stepped in, balancing two mugs of coffee. He spotted her immediately and pushed off the doorframe.

“Still here?” he said, handing her a cup. “I thought you left an hour ago.”

Elena wrapped her hands around the warm ceramic. “Just trying to get through these. It’s the same comments over and over. ‘Needs more evidence.’ ‘Unclear thesis.’ ‘Develop your analysis.’ I feel like a broken record.”

Marcus leaned against the counter, stirring sugar into his own coffee. “Yeah, grading’s the worst part. But hey—did you see the email from Reynolds? About the new AI teaching assistant?”

She frowned. “The one that’s supposed to help with lesson plans and feedback?”

“Exactly. District-mandated, starting next month. They’re calling it ‘EduAssist.’ It’s supposed to scan student work and generate draft comments based on rubrics. Saves hours, they say. I’ve been playing with it in my algebra classes—cuts my grading time in half.”

Elena stared into her coffee, watching the surface ripple. “And what does it actually say? Does it sound like a teacher wrote it?”

Marcus shrugged. “It’s not perfect. Sometimes the phrasing’s stiff. But you edit it before sending it back. Think of it as a first draft. You still get to make it yours.”

She thought of Javier’s essay, of the way his voice softened when he spoke in class, how his eyes lit up when he connected past struggles to present ones. Could a machine recognize that? Or would it just see a non-standard sentence structure and mark it down?

“I don’t know, Marcus,” she said quietly. “I don’t want to lose the part of teaching that’s… human.”

He nodded, sympathetic but earnest. “I get that. But what if it gave you back that time? What if you could use it to actually talk to kids instead of writing the same note twenty times?”

Elena looked down at the essay in front of her. Javier’s pencil had left a small smudge near the margin—a thumbprint, maybe, from when he’d paused to think. She imagined handing it back to him with a comment that didn’t just correct, but asked: What made you choose this example? The thought felt like a breath after holding it too long.

“Alright,” she said finally. “I’ll try it. Just for a week. See if it gives me any time back.”

Marcus grinned. “That’s the spirit. I’ll walk you through the login tomorrow. It’s easier than you think.”

She managed a small smile. “Thanks. For the coffee—and the nudge.”

As he left, the lounge door clicking shut behind him, Elena picked up her pen again. She didn’t write on Javier’s essay just yet. Instead, she turned to a fresh sheet of paper and began to list, in neat columns, the comments she’d written most often this semester. If the AI was going to take over the repetition, she wanted to know exactly what it would be learning to mimic. * The next morning, Elena arrived before the first bell, her laptop open on her desk. The EduAssist dashboard glowed softly on the screen—a clean interface with prompts for uploading essays and selecting feedback styles. She hesitated, then dragged Javier’s Civil Rights essay into the upload box.

The system whirred quietly. A progress bar filled. Then, in seconds, a comment appeared on the screen:

The thesis statement lacks clarity and specificity. Consider revising to clearly state the main argument and how it will be supported. Evidence is general; incorporate more specific examples from primary sources to strengthen analysis.

Elena read it slowly. It was accurate. It was also bland. She imagined Javier seeing it—would he feel seen, or corrected? She opened a new document and began to type, not replacing the AI’s words, but shaping them:

Your essay shows strong engagement with the Montgomery Bus Boycott, especially in how you connect Rosa Parks’ act to broader community organizing. To make your argument even clearer, try stating your main point upfront—what do you want the reader to understand about why the boycott succeeded? Then, use specific details, like the role of the Women’s Political Council or the carpool system, to show how that point played out in action.

She saved the edited version, then uploaded another essay. And another. The AI worked fast. Her job, she realized, was no longer to write from scratch—but to listen, to adjust, to make sure the feedback sounded like it came from someone who knew Javier’s quiet persistence, who had seen him stay after class to ask about primary sources, who believed, even when the data hesitated, that he was capable of more.

She leaned back, the morning sun now creeping across the floor. The stack of essays still loomed, but for the first time in weeks, it didn’t feel like a sentence. It felt like a beginning.

The following Monday, Elena arrived just after the first bell, her bag slung over one shoulder and a stack of student essays tucked under her arm. The hallway buzzed with lockers slamming and voices rising toward homeroom, but she slipped into her classroom, closed the door, and set her laptop on the desk. Sunlight pooled across the floor near the window, catching dust in the air like suspended thought.

She opened EduAssist and began uploading the latest batch—short analytical essays on the Civil Rights Movement, due Friday. One by one, the files disappeared into the upload box. The system responded with chimes and progress bars, then spat out feedback in uniform blocks of text.

The essay lacks a clear organizational structure. Ideas jump between events without transitions. Use chronological order or thematic grouping to improve flow.

Elena read it, then looked at the paper in front of her—Miguel Torres’s essay, which opened with a vivid vignette of his grandmother describing the 1963 March on Washington before looping back to discuss youth activism in Selma. The structure wasn’t linear, but it was intentional. The AI had mistaken creativity for confusion.

She opened a new comment box and typed: You begin with a powerful personal connection—your grandmother’s memory of the march—which draws the reader in. To help guide your audience through your thinking, consider using phrases like “This memory led me to wonder…” or “This made me ask how…” to link your opening to your analysis of Selma. Your voice is strong; now let your structure carry it forward.

She saved it, moved to the next.

The conclusion merely restates the introduction. Develop a deeper reflection on the significance of the events discussed.

Aisha Patel’s essay ended not with repetition, but with a quiet question: If the fight for justice requires both courage and care, how do we teach the next generation to balance them? The AI had missed the nuance, reading reflection as redundancy.

Elena revised: Your ending doesn’t just repeat—it invites us to think further. That’s rare and valuable. To build on that, try answering your own question briefly: what’s one way we might teach courage and care together? You’re already thinking like a historian—now let’s see where that question takes you.

By mid-morning, she had revised nearly every comment. The AI was fast—undeniably so—but its voice was flat, its gaze narrow. It saw deviations from the rubric as deficits, not differences. It didn’t know that Javier’s prepositional openings weren’t errors but rhythms, that Leila’s fragmented sentences in her diary-style entry were echoes of protest chants, that Marcus’s Algebra II student had written a proof in verse not because he couldn’t explain it, but because he wanted to sing it.

She paused, rubbing her temples. The screen blinked waiting for the next upload. Outside, a siren wailed in the distance, then faded.

Elena leaned forward and typed one last note—not to a student, but to herself, in the corner of the document: I am not here to correct the machine. I am here to teach the student. Let the AI handle the draft. I will handle the humanity.

She hit save, leaned back, and exhaled. The stack of essays still sat beside her laptop, but now, each one felt less like a task to finish and more like a conversation waiting to continue. * By third period, she had returned the first set of essays. Javier’s was among them. She watched from her desk as he unfolded the paper, his brow furrowed in concentration, then slowly smoothing as he read. He didn’t smile, but he nodded—once, sharply—before looking up. Their eyes met. In that glance, she saw not just recognition, but something quieter: the sense of being met halfway.

She didn’t know if the AI would ever learn to see what she saw. But for now, she was the one who saw. And that, she realized as the bell rang and students began to pack, was where the work began—not in the time saved, but in the moment returned.

The bell had just rung, scattering students into the hallway like loose pages caught in a draft. Elena remained at her desk, the laptop open to EduAssist’s lesson planner tab, the cursor blinking in the search bar. She had spent the morning grading, revising, returning—now, with a free period stretching ahead, she turned to the next task: planning the upcoming unit on the Harlem Renaissance.

She typed: Primary sources for 11th grade Harlem Renaissance unit, aligned to CCSS.ELA-LITERACY.RH.9-10.1 and RH.9-10.2. Within seconds, the AI returned a list: Hughes’s “The Negro Speaks of Rivers,” Hurston’s Their Eyes Were Watching God excerpt, a 1925 Opportunity magazine editorial, and a photograph of the Savoy Ballroom. Beside each, it noted the standard it addressed and suggested a discussion question: How does Hughes use river imagery to connect personal identity to ancestral history?

Elena leaned forward, scanning the output. It was efficient—too efficient, perhaps. The suggestions were accurate, standards-aligned, and drawn from reputable archives. But she knew her students. Last week, Miguel had connected the Great Migration to his family’s journey from Oaxaca. Aisha had written a poem in Spanglish about her grandmother’s kitchen as a site of resistance. Javier, quiet but incisive, had once compared the rent parties of Harlem to the fiestas in his aunt’s backyard in San Antonio—music, food, and storytelling as acts of communal survival.

She opened a new document and began to adapt. She kept Hughes’s poem but added a prompt: Choose a line that feels like a heartbeat. Now write your own line—about where you come from, or where you’re going. She swapped the Opportunity editorial for a letter from a Pullman porter to his wife in Chicago, found in the Schomburg Center’s digital archive—something tangible, personal. For the Savoy Ballroom photo, she added a short audio clip of Chick Webb’s orchestra, then asked: If this music had a color, what would it be? What would it feel like in your chest?

She paused, rereading the AI’s original question about Hughes’s imagery. It was sound, but it asked for analysis, not invitation. She revised it: What does this river carry besides water? What does it remember? Then, softer: Have you ever felt like you were flowing toward something you couldn’t name yet?

She saved the document, labeled it Harlem Renaissance – Day 1 – Draft, and leaned back. The screen dimmed slightly as the laptop dimmed its backlight. Outside, a gust of wind shook the oak outside her window, sending a shower of yellow leaves across the pavement.

She didn’t distrust the tool. She distrusted the assumption that efficiency equaled adequacy. The AI could gather, categorize, suggest—but it could not weigh the tremor in a student’s voice when they read aloud a line that mirrored their mother’s lullaby. It could not know that a question about ancestral rivers might land differently for a student whose family had crossed the Rio Grande, or the Atlantic, or the Sonoran Desert.

She reached for her notebook and sketched a quick timeline: Tuesday—Hughes and personal geography. Thursday—music and migration. Friday—student-created “rent party” presentations, where each would share a poem, song, or story rooted in their own heritage. She would use the AI’s list as a starting point, a scaffold. But the building? That was hers.

The door to her classroom creaked open. Marcus poked his head in, holding a styrofoam cup of coffee.

“You’re still here,” he said. “I thought you’d be gone by now.”

“I am,” she replied, smiling. “Just… finishing a thought.”

He stepped in, glancing at the screen. “Lesson planner treating you well?”

“It’s fast,” she said. “Like a really eager intern who’s read every book but never taught a class.”

Marcus chuckled. “Yeah, I’ve noticed it likes to default to the canon. Tried to give me a worksheet on quadratic formulas that sounded like it was written by a 1950s textbook.”

“Exactly,” she said. “It knows what’s standard. But not what’s alive.”

Marcus nodded, taking a sip. “So you’re tweaking it?”

“Always.” She closed the laptop gently. “The AI gives me the ingredients. I decide what to cook—and who I’m cooking for.”

He looked at her for a moment, then nodded again. “Well, whatever you’re making, it smells good. I’ll see you at lunch?”

“Definitely.” She gathered her things. “Thanks for the nudge yesterday. It’s helping—really.”

As she walked out, the hallway buzzing with lockers slamming and laughter, she felt the familiar weight of her bag against her shoulder—not just with papers and pens, but with the quiet responsibility of translation: taking what the machine offered and making it human. The work, she thought, wasn’t in the saving of time. It was in the choosing. And that, she realized, was where her attention belonged. Later that afternoon, during planning period, Elena returned to her desk with a fresh cup of tea. She opened EduAssist again, this time to generate a rubric for the upcoming Harlem Renaissance project. The AI produced a clean, four-category grid: Historical Accuracy, Analysis, Use of Evidence, and Presentation. Each had descriptors ranging from Exceeds Expectations to Below Standard.*

She read it slowly. It was clear, objective, easy to apply. But as she glanced at her notes—Miguel’s vignette, Aisha’s questioning closer, Javier’s prepositional rhythms—she frowned. Where was voice? Where was risk? Where was the space for a student to write something strange, beautiful, or uncertain?

She opened a new file and began to revise. Under Analysis, she added: Shows original thinking—connects ideas in ways that surprise or deepen the reader’s understanding. Under Use of Evidence, she wrote: Selects details not just for correctness, but for resonance—choosing moments that matter to the story being told. And under Presentation, she included: Demonstrates care in craft—whether through language, silence, rhythm, or form.

She saved it as Harlem Renaissance Project Rubric – Elena’s Version and printed a copy.

Elena leaned back in her chair, the printed rubric warm from the printer, and opened Javier’s essay again. She had saved a copy of EduAssist’s original feedback before her revisions—a habit she’d started to track patterns. The AI’s comments blinked on the screen in cool, neutral font: Your essay demonstrates solid historical understanding of the Montgomery Bus Boycott. However, your diction is frequently informal, which detracts from academic tone. Consider revising phrases like “the women got things moving” to more formal constructions such as “female activists played a pivotal role.” Avoid starting sentences with prepositional phrases, as this can weaken syntactic formality. Overall, work on elevating your language to meet the expectations of scholarly writing.

She read it twice. Her jaw tightened. Javier’s essay had opened with: In the heat of that summer, women like Jo Ann Robinson stayed up all night copying flyers by mimeograph machine. The sentence pulsed with rhythm and urgency—mirroring the clandestine energy of the boycott’s beginnings. The AI hadn’t just missed the power in his phrasing; it had labeled it deficient.

She remembered how Javier had sat forward during that unit, fingers tapping the desk when discussing Claudette Colvin’s arrest, how he’d asked, “Why don’t we hear about her more?” His voice had been quiet but insistent, not hesitant. She’d seen the same focus in his eyes when he read aloud Langston Hughes’s “I, Too” last week—his pronunciation deliberate, his pause before “I, too, am America” charged with meaning.

Elena opened another student’s essay at random—Aisha’s reflective piece on the March on Washington. Scrolled to the EduAssist feedback. There it was: Your use of rhetorical questions, while engaging, may be perceived as informal in an academic context. Consider rephrasing for declarative strength. Aisha had ended her essay with: What do we carry forward when the cameras turn off? The question hadn’t been a flaw—it had been the heart of it.

She pulled up Miguel’s vignette on the Great Migration. The AI had flagged his opening line: From the red clay of Georgia, my grandmother carried two things: a suitcase and a song. Comment: Avoid poetic openings in historical analysis; maintain objective tone from the outset.

Her breath came slower now. She wasn’t seeing isolated errors. She was seeing a pattern—one that punished linguistic rhythms rooted in oral tradition, in Spanish-influenced syntax, in the cadence of storytelling passed down through generations. The AI wasn’t assessing academic rigor; it was enforcing a narrow, monolingual ideal of what “scholarly” sounded like.

Elena opened Javier’s essay again. She clicked the settings menu in EduAssist, found the feedback toggle, and switched it off for his submissions only. A warning appeared: Disabling personalized feedback may reduce grading efficiency. She ignored it.

She leaned forward, began a new document. Not feedback. An audit. She listed each student’s name, then copied EduAssist’s original comment on diction or tone, followed by her own observation of the writing’s actual strength. Javier: AI said informal diction; observed—rhythmic urgency, prepositional phrases echoing Spanish syntax, powerful imagery. Aisha: AI said rhetorical questions too informal; observed—reflective depth, invitation to moral reckoning. Miguel: AI said poetic opening inappropriate; observed—vivid ancestral grounding, sensory immediacy.

She worked through half the class before the bell rang, the hallway noise swelling like a tide. She saved the file, closed the laptop, and sat still for a moment, hands resting on the cool wood of her desk.

Outside, the wind had died. The oak stood still, its last leaves clinging.

Elena stood, gathered her things, and walked toward the faculty lounge—not to complain, not yet, but to talk. To Marcus. To anyone who would listen. Because if the AI was measuring voice against a false standard, then efficiency wasn’t just inadequate—it was dangerous. And she had just seen the first thread in the pattern. * The faculty lounge smelled of burnt coffee and disinfectant. Marcus sat alone at a corner table, grading stack of worksheets, his brow furrowed.

“You look like you’ve seen a ghost,” he said, not looking up.

Elena slid into the chair opposite him, placed the printed audit on the table. “I’ve seen a bias,” she said.

He finally looked up, eyes flicking to the paper. “What is that?”

“It’s what EduAssist thinks is wrong with Javier’s writing. And Aisha’s. And Miguel’s. It’s not random. It’s systemic.”

Marcus picked up the first line—Javier’s essay—and read the AI’s comment. Then he read Elena’s note beside it. He went quiet. When he spoke, his voice was low. “You’re saying it’s not just missing nuance. It’s actively penalizing certain ways of speaking.”

“Exactly.” She tapped the page. “It’s not giving feedback. It’s enforcing a dialect. And it’s doing it in the name of academic rigor.”

Marcus ran a hand over his forehead. “I knew it liked outdated sources. I didn’t realize it was this… judgmental.”

Elena leaned forward. “I turned it off for Javier. Just for now. But I need to check everyone. I need to see how deep this goes.”

He nodded slowly. “You’re not just fixing feedback anymore. You’re auditing the algorithm.”

“I’m trying to see what it’s really teaching.” She paused. “And what it’s unteaching.”

Marcus closed his worksheet. “Then let’s do this right. I’ll pull the logs. We’ll look for patterns across classes. If it’s doing this to Javier, it’s probably doing it to others we haven’t noticed.”

Elena exhaled, the tension in her shoulders easing just a fraction. “Thank you,” she said. “I didn’t want to do this alone.”

He gave her a small, determined smile. “You’re not.”

She stood, gathered her things, and paused at the door. “I’ll see you after school?”

“Yeah. Bring the data. We’ll start with Javier—and go from there.”

As she stepped back into the hallway, the lockers shut with metallic clicks, the voices of students rising and falling like waves. She didn’t feel the weight of her bag so much as the weight of what she now carried: not just the responsibility to translate machine output into human meaning, but to question whether the machine had been taught to value the right things at all.

Elena waited outside Principal Reynolds’s office, the printed pages warm in her hands from being folded and refolded during her walk down the hall. The morning light slanted through the blinds, casting stripes across the linoleum. She had rehearsed what she would say, but now, with Javier’s essay in one folder and the AI’s feedback in another, the words felt too clinical. She wanted him to see—not just hear—what had been happening.

The door opened. Principal Reynolds looked up from his desk, where a half-empty coffee cup sat beside a stack of district memos. “Elena. You wanted to see me and Marcus?”

“Yes, thank you for seeing me on short notice.”

Marcus appeared in the doorway behind her, holding a tablet. He smiled, but it didn’t reach his eyes. “I brought the EduAssist logs. Just in case.”

They moved to the small conference room down the hall, the one with the whiteboard still half-covered in last week’s faculty meeting notes. Elena spread the papers across the table. Javier’s essay lay on top, his handwriting neat but urgent, the Montgomery Bus Boycott analyzed through the lens of grassroots organizing—Rosa Parks not as an isolated figure, but as part of a network of women who had laid the groundwork months before.

She pointed to the AI’s feedback, printed and highlighted. “It says here: ‘Needs improvement in academic tone. Avoid informal phrasing and ensure thesis is clearly stated.’ But look at this paragraph.” She tapped Javier’s opening line: ‘In the quiet hours before dawn, the women of Montgomery were already moving—mimeographing flyers, mapping routes, whispering names into telephones.’ “That’s not informal. That’s deliberate. He’s using a prepositional phrase to set a mood, to show the quiet labor behind the protest. It’s sophisticated.”

Marcus leaned in. “I see what you mean. The AI flagged ‘whispering names into telephones’ as colloquial?”

“Yes. And ‘mimeographing.’ It doesn’t recognize those as period-accurate verbs. It sees them as unprofessional because they’re not in its training data as formal academic terms. But Javier’s using them precisely to show how the movement worked—through covert, everyday actions.”

Principal Reynolds picked up the page. “So you’re saying the AI is punishing him for historical accuracy?”

“Not just that. It’s penalizing linguistic patterns common in bilingual writers. Starting sentences with prepositional phrases—like Javier does—is a feature of Spanish-influenced English. The AI reads it as fragmented or immature, when it’s actually a stylistic strength. I’ve seen it flag Aisha’s reflective questions and Miguel’s vignettes the same way.”

Marcus ran a hand over his jaw. “I thought the tool was neutral. That it was just matching language to rubrics. But if it’s penalizing features that aren’t errors… then it’s not neutral. It’s biased.”

Elena nodded. “It’s not grading writing. It’s grading conformity to a narrow model of academic English—one that excludes the very students we’re trying to uplift.”

Principal Reynolds set the paper down slowly. His gaze drifted to the window, where a group of students laughed near the bike rack. “I pushed for this because the district said it would save time. Reduce workload. But if it’s doing this… then we’re not saving time. We’re reinforcing inequity.”

Marcus looked at Elena. “I didn’t see it until now. I trusted the outputs because they seemed consistent. But consistency isn’t fairness if the standard is flawed.”

Elena felt the tightness in her chest ease, just a fraction. She hadn’t come to accuse. She’d come to show. And now, seeing Marcus’s troubled expression, she knew the evidence had landed—not as an argument, but as an invitation to look closer.

Later, after Principal Reynolds had gone to call a meeting with the tech committee and Marcus had returned to his classroom to recheck his own EduAssist reports, Elena stayed behind. She stacked the papers neatly, then slipped Javier’s essay back into her bag. On the table, she left a single sticky note with her own words from the night before: I am not here to correct the machine. I am here to teach the student. She didn’t need to say more. The paper would speak for itself when others picked it up.

The faculty lounge smelled of stale coffee and citrus cleaner, the hum of the refrigerator a constant backdrop to the low murmur of voices. Sunlight filtered through the blinds, casting alternating bars of light and shadow across the worn linoleum. Chairs were pulled into a rough circle, the whiteboard at the front blank but for a single marker resting in its tray. Teachers drifted in—Ms. Delgado from English, Mr. Patel from science, Coach Ruiz—each carrying a mug or a stack of papers, settling with the quiet resignation of those who’d attended too many mandatory trainings.

Elena stood by the window, watching a sparrow hop across the fire escape, until Principal Reynolds cleared his throat. “Thanks for coming, everyone. I know it’s short notice. Elena’s got something she wants to share.”

She stepped forward, the sticky note from the conference room still tucked in her pocket. “I’m not here to tell you EduAssist is useless. I’ve seen how it can help—suggesting sources, catching basic errors, giving us a starting point. But I’ve also seen how it misses the mark. Not because it’s broken, but because it’s learning from a narrow idea of what ‘good’ writing looks like.”

She unfolded Javier’s essay and laid it on the table, then placed the AI’s feedback beside it. “This is Javier’s opening line: ‘In the quiet hours before dawn, the women of Montgomery were already moving—mimeographing flyers, mapping routes, whispering names into telephones.’ The AI called it informal. But it’s not. It’s precise. It’s historical. It’s the kind of writing we want to encourage—writing that shows understanding, not just grammar.”

Ms. Delgado leaned forward. “So it’s not just about tone. It’s about whose language counts as academic.”

“Exactly,” Elena said. “And it’s not just Javier. I’ve seen it flag Aisha’s reflective questions, Miguel’s vignettes—anything that doesn’t sound like a textbook. If we let it give feedback without checking, we’re telling students their voices don’t belong here.”

Mr. Patel crossed his arms. “But we’re supposed to save time. If we have to review every suggestion, what’s the point?”

“The point,” Elena said, “isn’t to replace our judgment. It’s to free us up to use it better. The AI can draft. We revise. We add the context—like how Javier connected the boycott to the Women’s Political Council, or how he used period-accurate verbs to show the covert work behind the protest. That’s not in the AI’s training data. That’s ours.”

A pause. Coach Ruiz rubbed the back of his neck. “So we use it, but we don’t trust it blindly.”

“We guide it,” Elena said. “And we tell students when it’s involved. Transparency matters. If a student gets feedback shaped by an AI, they should know—so they can question it, learn from it, not just accept it as authority.”

Principal Reynolds picked up the AI’s feedback sheet. “What if we made this official? Not a ban. A framework.”

Elena nodded. “Three things: First, teacher oversight. AI feedback is a draft, not a final word. Second, bias review. We check outputs against our knowledge of students—especially those whose language or style might be misread. Third, student transparency. Let them know when AI is used in their work, and invite them to reflect on it.”

Mr. Patel tapped his pen. “Could we make that a rubric addendum? A quick checklist?”

“Yes,” Elena said. “And we could share examples—like this one—so we’re not starting from scratch.”

Ms. Delgado smiled. “I’ll bring it to the English department. We’ve been talking about this for months.”

The discussion flowed then—practical, grounded. They sketched a one-page guideline on the whiteboard: AI as Supplement, Not Substitute. Review for Bias. Inform Students. They talked about how to build time for review into grading routines, how to flag problematic patterns in the AI’s logs, how to create a shared folder of revised feedback samples.

When the bell rang, signaling the end of lunch, no one rushed to leave. Instead, they lingered, copying the points onto paper, taking photos with their phones.

Elena stayed behind, erasing the whiteboard slowly. Principal Reynolds appeared in the doorway, holding two cups of coffee.

“You okay?” he asked.

She took the cup, the warmth seeping into her palms. “Yeah. I think we’re going to be all right.”

He leaned against the doorframe. “I used to think efficiency was the goal. Now I see it’s not about saving time—it’s about what we do with the time we get back.”

She looked at him. “Exactly. The work isn’t in the grading. It’s in the moment we return to the student.”

He nodded, then paused. “That sticky note you left? I showed it to the tech committee. They’re going to print it and put it next to every EduAssist login screen.”

Elena let out a breath she hadn’t realized she was holding. “Good.”

He smiled faintly. “Welcome to the future, Elena. Hope it’s a little more human than we thought.”

She turned back to the window. The sparrow was gone. In its place, a breeze stirred the leaves of the sycamore outside, scattering light across the pavement. She took a sip of coffee, bitter and real, and for the first time in weeks, the stack of essays on her desk didn’t feel like a burden.

It felt like a beginning.