Top 10 AI News This Week — November 2–8, 2026
Nov 9, 2026
The Education Department proposed writing the words “artificial intelligence” out of every one of its grant priorities, which means the language your next federal proposal is graded against is about to change. The week's two strongest items, though, are audits rather than announcements: a community college professor found that every published meta-analysis claiming AI improves learning fails basic standards, and a vendor vulnerability cost about 400 California community college students their financial aid refunds. Here's what community college educators should know.
EdTech & Campus AI
The Education Department is writing “Super Intelligence” into every grant priority — and comments close December 7
On November 5 the Department of Education published Proposed Priorities and Definitions—Secretary's Supplemental Priorities and Definitions at 91 FR 69736, FR Doc. 2026-22615, docket ED-2026-OS-3169. It proposes, verbatim, “to use the term ‘Super Intelligence’ to replace ‘Artificial Intelligence’” and to “use ‘SI’ to replace ‘AI’ in every use of the terms.” The proposed definition is a pointer, not an expansion: “Super intelligence (SI) means the technologies and systems encompassed by the term ‘artificial intelligence’ as defined in 15 U.S.C. 9401(3).”
It amends four existing priorities, not one: Evidence-Based Literacy, Education Choice, and Returning Education to the States (90 FR 43514); Meaningful Learning Opportunities (91 FR 6625); Career Pathways and Workforce Readiness (91 FR 18780); and Advancing Artificial Intelligence in Education (91 FR 18774). Comments are due “on or before December 7, 2026,” through Regulations.gov only — “ED will not accept comments by fax or by email, or comments submitted after the comment period closes.”
Here is why this is more than vocabulary. Executive Order 14434, which we covered last week, exempts grants and contracts from its renaming mandate. ED is extending it to its grant priorities anyway, ahead of the proposed statutory definition due from the White House around November 28. One flag: the document is internally inconsistent, citing E.O. 14434 in its rationale and E.O. 14344 in its definition. Add the typo to your search strings, because agency documents will carry it forward.
For community colleges, this is concrete: every future ED discretionary competition Fullerton and NOCCCD enter will be written in SI vocabulary, and grant alerts, proposal boilerplate and internal templates keyed to “AI” will quietly miss opportunities. The comment window is the cheapest policy action available to a college this cycle. Read more →
Hackers redirected financial aid refunds at 90 California community colleges — and there is no AI in this one
We are running this because financial-aid integrity is a beat we track and because it is the most consequential thing on a California community college administrator's desk this week. It is not an AI story, and we are not going to pretend otherwise: no AI was involved in the attack.
Reported November 5 by Abby Sourwine for Government Technology. BM Technologies, formerly BankMobile, changed its account-verification procedure on July 1. Attackers altered identity-verification data and, for roughly 400 students, changed bank and routing information to divert refunds. BM Technologies estimated that data changes touched about 65,000 student accounts, with intruders accessing about 1,000. 56 districts, covering 90 colleges, use the vendor to distribute refunds. The Chancellor's Office security center noticed the problem in early October and told colleges to pause disbursements through the vendor.
Chris Ferguson, executive vice chancellor of finance and strategic initiatives, was careful about the framing: “We want to emphasize that this was not a breach,” and “This was a bad actor taking advantage of a vulnerability that allowed them to take over certain accounts.” Rancho Santiago CCD moved more than 3,000 student payments to paper checks. Los Rios switched to Nelnet; spokesman Gabe Ross: “Moving to Nelnet addresses future disbursements, but we remain focused on students with unresolved BankMobile issues.”
A sourcing note: EdSource broke this on October 30, outside our window, which is why it appears here now rather than last week. If you work in a business office or a financial aid office, the action item is the dullest possible one — confirm which vendor your district uses for refund disbursement, and whether anything is paused. Read more →
A community college professor audited the research claiming AI improves learning. All 14 studies failed.
Education Week, November 2 (Stephen Sawchuk). Patrick O'Neill, associate professor at Ivy Tech Community College in Bloomington, Indiana, audited 14 of the nearly 40 published meta-analyses reporting positive AI effects on learning. In an EdArXiv preprint, he reports that all 14 failed basic methodological standards: unclear definitions of the AI use being studied, weak construct validity, inconsistent outcome measures, and no control for publication bias, with all but two failing to control for dependent variables. Some of the underlying studies used no AI tools at all; some carried impossible statistics or likely hallucinated references.
Those 14 papers appeared in 11 well-regarded journals and have been cited more than 3,000 times. One, published in a Nature journal in May 2025, was retracted a year later. O'Neill's diagnosis is about incentives: “There is a hunger for this evidence which is driving people to publish quickly and overstate the solutions.” Independent corroboration of the scale of the problem: Stanford's AI Hub for Learning reviewed more than 800 studies for its 2026 report and found 20 with strong causal design. Hechinger's Jill Barshay, who broke the story on October 27, adds that the field is generating roughly 200 new citations a month, that two-thirds of AI-in-education calls for papers specifically invited research on benefits, and that O'Neill's own preliminary meta-analysis of 44 well-designed experiments averaged roughly zero effect.
We could not read the preprint directly — OSF and EdArXiv pages return front matter only to automated retrieval — so the findings above are as reported by EdWeek and Hechinger. If you teach: the next time a vendor deck or a dean's memo cites a meta-analysis showing AI learning gains, the odds are better than even that it is one of these. If you sign purchase orders or set professional-development budgets, the honest state of the evidence is that there is no defensible basis right now for a generative-AI learning-gains claim. The person who established that teaches at a community college. Read more →
A tutorial finally prints the per-clip bill: about $16 to $24 for a finished AI video
Our standing rule on creative-tool coverage is that a tutorial which states no prices is not costed advice, however good the technique. Curious Refuge published four posts in the window and two of them price something.
“How to Create a Realistic AI Clone,” November 5, is the properly costed one and the first on this beat to give a per-project figure from a practitioner rather than a vendor: clean the audio in Adobe Podcast, export over black video in Premiere, pair it with a short reference clip, and generate in Seedance 2.5. Stated costs, explicitly as per-example figures: about $16 for Seedance 2.5 at 1080p, $0.05 for HeyGen, and $24 for the fal.ai-trained LoRA workflow. “How to Remove Music from Your Videos Using AI,” November 6, uses Meta's SAM Audio, which the post says is completely free, to strip a music bed while keeping dialogue and effects, and includes recovery steps for when separation damages sound you wanted to keep.
The other two posts, both dated November 2, are the “Best AI Video Generators for 2027” and “Best AI Image Generators for 2027” roundups. We are not running either as advice. Neither prices anything, and in the video roundup the published overall scores do not reconcile with the category averages the same page prints. The image roundup discloses more method — 29 standardized edge-case prompts scored 1 to 10 across three categories, against a real reference image — and its arithmetic does check out, so it is usable as a test harness you could reuse, not as a ranking to trust.
For film, television, digital media arts, graphic design and marketing faculty, the $16-to-$24 range is the number that belongs in a project brief before students start generating, and it is the first time we have been able to publish one. The SAM Audio workflow has a second, duller use worth knowing: it is a practical way to reuse lecture or event footage that has uncleared music under it. Read more →
Major AI Companies
Microsoft is building an IT lock that pins students to the tutor — and quietly capping Copilot for unlicensed users
The do-something-this-week item, and the webinars are tomorrow. Microsoft's education blog, November 2 (updated November 3), announces a forthcoming mode described verbatim as “An IT controlled locked down learning-only mode that always lands students in Study and Learn.” In it, students cannot close the agent, switch to general Copilot Chat, or reach other agents or features; they can still attach their own files as references. Planned coverage includes Copilot Chat on web and app, the agent store, direct links and previous chats. Four webinars run November 10 and November 12. This is announced, not shipped, and no GA date was given.
The baseline matters as much as the announcement. The Study and Learn Agent is available to Education users aged 13 and over with a Microsoft 365 Education A1, A3 or A5 license, and “No additional license or purchase is required.” Pre-selection is on by default for K-12 tenants and off by default for higher education — so a community college starts with this switched off and has to opt in deliberately.
Separately, and with a caveat we are not going to paper over: message center items MC1536158 and MC1536163, published November 3, say Microsoft is extending “fair use” limits to further Copilot Chat features for users without a Microsoft 365 Copilot license in commercial tenants, with rollout beginning November 9. Access “can vary throughout the day”; users are notified in-product with a retry timeframe; in-flight responses are not interrupted; and Microsoft sets and manages the limits, with admins unable to view or adjust them. No limit figures are published anywhere.
Whether this reaches colleges is genuinely unresolved. Neither message names education SKUs, and Microsoft's own Copilot Chat requirements documentation, refreshed November 5, still does not say whether A1, A3 and A5 count as “licensed” for this purpose. Most community colleges run Copilot Chat on those licenses without the paid add-on — which is exactly the population described, if education tenants are in scope. We are reporting the ambiguity, not asserting an answer. For IT staff, the diagnostic is aka.ms/CopilotLicenseDetails. Read more →
Google put Guided Learning in Classroom worldwide, and funded a California community college to teach teachers
Three Google threads, November 3 to 5, and one of them is local.
Guided Learning is now assignable in Google Classroom globally, on Education Fundamentals, Standard and Plus — editions community colleges actually hold, so this is not a K-12-only release. The admin gate is that teachers and students must sit in a group or OU with Gemini set to On. Teachers can see AI-generated insight summaries and individual student chat transcripts, which extends the teacher-visible-transcript question we raised on October 19 into a second surface and is a FERPA and syllabus-disclosure matter, not a feature note. Google also said access to Gemini, Gemini Notebook, Gemini in Classroom and Google Search “will be entirely off for K-12 student accounts until school administrators explicitly turn them on” — that change is K-12 only and does not reach college tenants.
Second, Google.org announced “over $9.5 million” across five recipients: the City University of New York, Fab AI, the Krause Center for Innovation, Teaching Lab and 1EdTech Global. No per-organization amounts, and no application process, criteria or deadline — this is an announcement of money already placed. The reason it matters here: the Krause Center for Innovation is housed at Foothill College, in the Foothill-De Anza Community College District, which makes a California community college one of five named recipients of a national Google.org AI-education commitment. Google's own post names only the center, not the college; we verified the Foothill affiliation separately from the center's own site.
Third, the paid tiers moved and most colleges get nothing. Gemini in Google Forms is Education Plus plus the Google AI Pro for Education or Teaching and Learning add-ons, with Scheduled Release on November 19 — the date for admins to note. Gemini Omni 1.2 Flash in Google Vids, generating 15-second clips, is gated the same way, and Google states plainly that “This feature does not have an admin control.” If your college is on Fundamentals or Standard, neither is yours. Read more →
Two dates we put on your calendar: one lands Thursday, the other is gone
Anthropic's revised Usage Policy takes effect Thursday, November 12. There was no announcement in window on either of Anthropic's newsrooms; the live policy simply carries the date. What it does to college business is specific. Recommending the admission, rejection or ranking of applicants is a High-Risk AI Recommendation. So is grading an exam that determines admission, graduation, licensure or certification (multiple-choice testing is excluded), making academic misconduct determinations, and suspending or expelling students. For any of those, a qualified person must review the output before it is used and the affected individual must be told AI was involved. Explaining general educational content is excluded. The fraud provisions bar submitting AI-assisted work “without proper permission or attribution.” “Minor” means under 18.
If any part of your admissions, grading-for-licensure, integrity or discipline process touches Claude — nursing and EMT certification exams, CTE certificate assessments, an academic integrity committee — the human-review-and-disclosure requirement attaches in three days.
And Meta's November 20 deadline is off. On November 5, per Nieman Lab (Hanaa' Tameez and Andrew Deck), Meta removed the preview link to its proposed terms and posted a Business Help Center update pausing four documents indefinitely: the Meta Commercial Terms, Product Catalog Terms, Self-Serve Ad Terms and Ad Creative Generative AI Terms. A spokesperson: “We're pausing the changes as we continue to engage with stakeholders and take our time to get this right.” Existing terms stay in force. That is the second slip — October 30, then November 20, now paused.
We told readers on October 19 and again last week that November 20 was a date for marketing offices, athletics and foundations running Meta ads, because the contested clause would have let Meta crawl websites associated with an advertiser's products. Stand down on that review. One honesty note: there is no primary source for the pause. Meta's own news site carries nothing, and its legal pages are closed to automated retrieval, so this rests on reporting. Read more →
Free & Student Opportunities
The Google educator badge is real, it shipped, and it stacks into college credit
A question this digest has been carrying for weeks is answered, and the answer is yes. We have reported three times that educertifications.google is JavaScript-gated and would not confirm whether the Google AI Educator Series actually issues a badge. It does. Confirmed on the readable mirror and Google's own support FAQ: a digital certificate badge per session, earned by passing a short quiz, issued through Accredible, appearing 5 to 20 minutes after you pass.
Cost, stated honestly: genuinely free, with no audit-versus-certificate split and no upsell we could find. The blog calls it “a no-cost, on-demand training” and the FAQ says “free of charge for every educator that takes part,” with paid Gemini features flagged where a module uses them. November's modules went live November 4 in the US: building interactive assessments, game-based learning, and strengthening academic assignments. Higher-education modules run 30 to 45 minutes, on demand, no prerequisites, no fixed order, with new content the first Wednesday of each month.
The genuinely new part, and the reason this leads the section: the FAQ says badges can be stacked into accredited undergraduate or graduate college credit, or CEUs, through College Unbound, ISTE with Dominican University, and the Illinois Digital Educators Alliance. For faculty on a salary schedule that rewards units, that converts a free 45-minute module into something with a dollar value. One fair caveat raised by critics of the series: it teaches educators to use AI, not to teach AI literacy to students.
Beyond that, nothing launched, and that is the finding rather than a gap. Anthropic's catalog now carries AI Fluency for Creative Work and Introduction to Claude Cowork, both free with a certificate — but Anthropic publishes no dates anywhere on that catalog, so we cannot say they are new and we are not claiming it. OpenAI Academy's catalog was last touched around October 27; all 14 courses are free with a badge at 80% on the assessment, and OpenAI says plainly they are “not certifications.” AWS Training and Certification: nothing since October 27. CVC@ONE: no new cohort. Coursera and edX: nothing genuinely free — and note the trap, since HP's “AI in Society” on edX is free to audit but “certificates are only available for paying learners.”
One competing item, out of window and flagged as such: Microsoft's GitHub Copilot Pro Badge went GA October 28. It is free, needs no exam, accrues from real Copilot usage across github.com, the IDE, CLI and mobile, and is claimed on Credly. The blocker for us is specific: Enterprise Managed User accounts are unsupported at launch, which is how many colleges issue GitHub, so students would need personal accounts to earn it. Read more →
No new grant deadline a college could chase — and one eligibility line to read before you plan
Both NSF AI solicitations closed November 4 and nothing new was posted. NSF 26-513 was last updated October 28; NSF 26-512 shows a November 3 timestamp with no revision note and no change to the deadline, which is worth saying because an NSF timestamp is not a change indicator. Both recur on the first Wednesday in November, so the next deadline is November 3, 2027. Eligibility remains verbatim “Two- and four-year IHEs (including community colleges)” that are “accredited in, and having a campus located in the U.S.,” “acting on behalf of their faculty members.” But read the limits before planning a lead application: 26-513 states that only one award per state or multi-state region will be made, with one proposal per organization and one per principal investigator. For a California community college that argues for joining a consortium, not leading one.
The California Workforce Development Board's High Road to Construction Careers Round Two carries roughly $19 million for this state fiscal year and has no posted deadline. Before anyone builds a plan on it, read the eligibility line on grants.ca.gov, which lists eligible applicants as “Business, Individual, Nonprofit, and Tribal Government” — public educational institutions are not on that list as direct applicants — and note that Round One was open to existing HRCC regions only. A college's route here is as a partner, not an applicant.
The Gates AI Transformation Learning Grants site launched November 2, but concept papers closed September 11 and notifications went out in October, and the eight recipients still have not appeared in the committed-grants database, so there is no citable list yet. Eligibility was verbatim “Broad-access higher education institutions—including community colleges, public regional universities.” We will keep checking; a California name is plausible.
So the honest summary is that no new fundable deadline appeared this week. The one open federal window is not money at all: comments on the Education Department's proposed priorities close December 7, which is the first story in this issue. Our AI scholarship check is a first-run-of-the-month task and ran on November 2; the next one is December 7. Read more →
AI Competitions
Four competitions closing in eleven days, and the biggest one understates its own prize money
Every eligibility, deadline and prize figure below was read off the live main-page sidebar, never the rules tab, and every deadline is converted to Pacific — which since November 1 means PST, so recheck any sidebar that still says PDT.
PayPal AI Hackathon — Thursday, November 12, 12:00pm PST. Eligibility verbatim: “Above legal age of majority in country of residence” and “All countries/territories, excluding standard exceptions.” No student restriction, so faculty and staff can enter; anyone under 18, including dual-enrolled high schoolers, cannot. Free. 31,860 participants, the largest open AI hackathon we found. And here the marketing runs the other way for once: the page advertises “$67,500+ in prizes,” and $67,500 of it is cash across 17 awards topping out at $12,000, summing exactly — with $2,250 in Render credits and a six-month APIMatic subscription on top, which is where the page's “$69,750 in total prizes” comes from. Given this section's track record, an advertised figure that understates the cash deserves to be said out loud. Enter here
UGREEN Hackathon 2026 on AI NAS — Round 1 Sunday, November 15, 11:45pm PST (finalists' Round 2 November 30). 18 and over, “Companies/professional organizations excluded,” some countries excluded but the US is eligible. Free, 1,781 participants. $46,000, all cash — $20,000, $10,000, $6,000 and two $5,000 Most Creative awards, summing exactly. Faculty may enter as individuals.
NASA ORBIT Challenge 2027 — registration closes Monday, November 16, and this is the one built for a faculty member. NASA's page says the challenge is for “university and community college students nationwide,” and the rules add hard structure: a Lead Institution on the FY2026 Eligibility Matrix, and a faculty advisor who is an employee of that institution and is the team lead for eligibility purposes. Teams run up to six undergraduate or graduate students, at least 60% from the lead institution, each 18 or over and a US citizen, lawful permanent resident or qualifying foreign national. Up to $500,000, every dollar of it dollar-denominated with no credits — but read this before you pitch it to students: “Prize payments go to the Lead Institution,” not to the students, and Phase 1 carries no purse at all. Phase 1 submissions close December 14, 3:00pm ET. NASA publishes two different closing dates; the November 16 registration cut-off is the operative one.
Meta VR Start Developer Competition 2026 — Wednesday, November 18, 12:00pm PST. Re-verified this week and the figures hold: $1,000,000, with all 19 line items cash, summing exactly, no credits, no licences, no hardware. 18 and over, no student restriction, free, and participation rose from 5,470 to 5,528 during our check. It remains the only seven-figure all-cash entry on the board.
One just past the window that is worth starting now: Kiroween, Friday, December 4, 2:00pm PST — $100,000, all cash, summing exactly, with four “Best Student Project” awards of $1,500 cash where all team members must be students. 10,767 participants, free, no student restriction on the main categories. Italy and Quebec are among the excluded territories; the US is eligible.
“$28,000 in crypto” that pays nothing, a hackathon that contradicts itself, and three corrections we owe
Ctrl+V Hackathon advertises “$28,000 in crypto,” cryptocurrency icons and all. Every prize is Adaption Platform credits — three at $1,000 and a hundred at $250. Actual cash: zero. It closes November 10, so it is effectively gone, but the framing is why it is here.
IceHacks (Sunday, November 15, 6:00pm PST) advertises “$17,150+ in prizes” and the same page states “IceHacks doesn't award any cash prizes.” Both sentences are live simultaneously. The listed figures do sum to $17,150, but every tier is bundled with vendor credits and the organiser says in writing that it pays no cash, so treat the headline number as unproven. Eligibility: ages 13 and over, students only, so faculty and staff cannot enter.
Microsoft Imagine Cup 2027 opened for the new season, closing February 28, 2027, 11:59pm CST. Founders do not need to be students, so faculty and staff are eligible, and there is a Student Startup lane. The honest math: $350,000 in cash and $1,950,000 in startup credits. The champion gets $250,000 cash plus $1,000,000 in credits; the Top Student Startup gets $25,000 cash plus $250,000 in credits. Any “$2.3 million” framing is about 85% vendor credit. Health Hackathon 3.0 advertises “$200,000 in cash” while its listed amounts sum to 190,000 with no currency symbol — and the US is not among its included countries, so Californians cannot enter it anyway.
Three corrections. First, STEM2CONNECT has four non-cash prizes, not three. We published three. The fourth is a free first-year .xyz domain to 44 teams that submit on time, first come first served. The November 21 deadline and the zero-cash finding both stand. Second, Codex x TwitchCon is closed. We ran it as a $25,000 all-cash item with a November 2 deadline; it no longer appears anywhere in the live open query, so do not treat it as available. Third, Kaggle and HackerEarth were unreadable again — both listing pages are JavaScript-rendered and return only page metadata. We ran nothing from either and, per our own rule, inferred no dates from past cycles. Read more →
Embodied AI & Robotics
Robot production is taking over car factory floor space — and one humanoid company now has to file a 10-K
Agility Robotics began trading on Nasdaq on November 5 as AGLT (warrants AGLTW), the first US publicly listed pure-play humanoid robotics company, after completing its business combination with Churchill Capital Corp XI on November 4. About $465 million gross before transaction expenses, including a PIPE led by Foxconn and $100 million raised before closing. Customers already running Digit: GXO, Schaeffler and Toyota Motor Manufacturing Canada, with AGCO signed for a 2027 pilot and three unnamed enterprise manufacturers in paid programs. More than $300 million in committed multi-year Digit 5 orders, subject to milestones. Nearly three years of real-world deployment and more than 65,000 operating hours; at GXO's Georgia site Digit has moved more than 100,000 totes, with two more US GXO facilities coming in early 2027. Digit 5 lifts up to 50 pounds repetitively, recharges in about nine minutes, and is claimed to deliver more than 20 productive hours a day. Facilities: RoboFab in Salem, Oregon, 70,000 square feet, designed for up to 10,000 robots a year; a 50,000-square-foot Pittsburgh expansion; and hardware and physical-AI development in Fremont, California.
Pointing the same direction from the other end: on November 6 Tesla said it will end Cybertruck production in 2027 and convert the floor space to the two-seat Cybercab. The sourcing deserves to be stated plainly — this is a Musk post on X, reported by TechCrunch, not a filing and not an earnings call. Musk once projected roughly 250,000 Cybertrucks a year; actual sales are in the tens of thousands across three years. No Cybercab volume figures were given. It follows the already-completed decision to end Model S and Model X production and hand that space to Optimus.
For automotive technology programs, the Tesla half is the sharper one: a nameplate students were being trained to service is being retired, and the line is retooling for a vehicle with no steering wheel, which shifts a service curriculum toward sensor calibration, high-voltage work and teleoperation support. For robotics, mechatronics and logistics faculty, the Agility listing is the more useful event, and the reason is unglamorous. A pure-play humanoid company now has to file 10-Ks and 10-Qs. That means audited unit economics, real deployment hours and an order backlog you can read, instead of demo videos — a free, classroom-ready antidote to the hype problem that makes this beat so hard to teach from. Note what the release does not contain: no headcount, no technician counts, no training program. Read more →
Drone delivery raised $1.5 billion, promised 100 factory jobs, and ran into the neighbors
Three items that only make sense together. Zipline, November 5: a $1.5 billion round at a $21.5 billion valuation, up from $7.6 billion in January 2026, led by Paradigm, WCM Investment Management and Layer Global, alongside a new multiyear Walmart agreement adding 300 sites. More than a million deliveries in Dallas–Fort Worth during 2026; 150 million commercial autonomous miles with no safety incident; an October median delivery time of 19 minutes from checkout. Proceeds go to manufacturing, hiring and delivery infrastructure — with no hiring numbers, no training detail and no California sites named, which is why it runs as a market-scale signal and not a campus story.
Wing, November 6, supplies the number this beat usually lacks: up to $125 million in direct US manufacturing over two years, a 15,000-square-foot Fort Worth plant with claimed capacity of tens of thousands of aircraft a year and up to 100 manufacturing jobs, serving Walmart across 270-plus stores by 2027. Wing ties the expansion to the pending Part 108 rule, in White House review since July 10 against a December 7, 2026 statutory deadline. DroneXL's scepticism belongs in print alongside it: Wing has not said how many aircraft fly its 70-plus existing sites, and the Dallas Morning News reports the Fort Worth operation actually opened earlier this year. Wing also referenced California research and engineering spending inside that $125 million without breaking it out, which we will chase.
And then the constraint, from IEEE Spectrum on November 4. No decibel figures — residents in Baton Rouge counted a drone every 48 seconds on a Saturday morning and compared the rotor whine to a weed whacker. Amazon's MK30 weighs about 36 kilograms and must fly low; an FAA environmental assessment projected 1,000 daily operations, some between 6am and 10:30pm. On September 28, fifteen states and Harris County, Texas sued the FAA over eased environmental review for drone sites; Cobb County, Georgia denied a Wing zoning request in April; Westborough, Massachusetts voters approved commercial drone restrictions.
If you teach UAS or drone technology — and Fullerton does — this is the week's assignable reading. Programs teach airframes, Part 107 and payloads; the binding constraint on the industry your students are entering is now zoning, community acceptance and environmental review. The Spectrum piece works as-is, and it pairs directly with the Wing and Zipline announcements as the thing those announcements have to get past. Read more →
This roundup was assembled with AI-assisted research by the EDUAIATLAS team. Spot something we missed? Let us know.