# References and Further Reading

UNC Water & Health 2026 · Side Session · Baobab Tech · October 2026

Numbers match the citations on the slides, in order of first appearance. Research notes link to the evidence summaries in [baobab-tech/research-ai-energy](https://github.com/baobab-tech/research-ai-energy).

## References

| # | Reference | Supports |
|---|---|---|
| 1 | Luccioni, A.S., Gamazaychikov, B., Alves da Costa, T. & Strubell, E. (2025). *Misinformation by Omission: The Need for More Environmental Transparency in AI.* arXiv:2506.15572. https://arxiv.org/abs/2506.15572 | In May 2025, 84% of LLM token usage on OpenRouter's top 20 models went through models with no environmental disclosure |
| 2 | Elsworth, C. et al. (Google) (2025). *Measuring the environmental impact of delivering AI at Google scale.* arXiv:2508.15734. https://arxiv.org/abs/2508.15734 | Median Gemini text prompt: 0.26 mL on-site cooling water (scope 1) |
| 3 | Li, P., Yang, J., Islam, M.A. & Ren, S. (2023, rev. 2025). *Making AI Less "Thirsty".* arXiv:2304.03271. https://arxiv.org/abs/2304.03271 | 2.2 mL per request on site (scope 1), 16.9 mL including electricity (US average); GPT-3 training 5.44 million litres, 87% from electricity generation |
| 4 | Akinade, Amanambu, Frame & Ren (2026). arXiv:2606.21760. https://arxiv.org/abs/2606.21760. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/water/033-akinade-2026-water-consumption-impact-utility-burden.md) | Meta, Lebanon, Indiana: peak-day demand 1.34x the host utility's capacity. **Caveat:** that is the current municipal system; it falls to 0.25x against the dedicated supply being built, and the peaking factor is modelled, not metered. |
| 5 | Congressional Research Service (31 Jul 2026). *Data Centers and Water* (R49057). https://www.congress.gov/crs-product/R49057. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/community/026-crs-2026-data-centers-water-municipal-supply.md) | Richland Parish: permitted 8.4 bn gal/yr against 0.5 to 0.6 bn stated (about 14x); rates and service agreements often confidential |
| 6 | Han, Y., Wu, Z., Li, P., Wierman, A. & Ren, S. (2026). arXiv:2412.06288 (v4). https://arxiv.org/abs/2412.06288. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/community/023-han-2026-data-center-health-costs-county.md) | $20.9bn and 1,262 deaths in 2028 under high growth; over 90% from generating plants |
| 7 | Luccioni, A.S., Jernite, Y. & Strubell, E. (2024). "Power Hungry Processing." FAccT '24. https://doi.org/10.1145/3630106.3658542. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/frugal/027-luccioni-2024-power-hungry-processing.md) | Image generation 2.907 kWh per 1,000 inferences against 0.002 kWh for text classification (>1,450x); extractive QA: task-specific 0.3 gCO₂e against 10 g for a multi-purpose model |
| 8 | Morrison, J., Smith, N.A. & Strubell, E. (2026). arXiv:2605.01158. https://arxiv.org/abs/2605.01158. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/frugal/018-morrison-2026-olmo3-reasoning-posttraining-17x.md) | Reasoning post-training cost 17x the instruction variant (OLMo 3) |
| 9 | Roussilhe, G., Dromard, B. & Mitra, S. (2026). arXiv:2604.26539 (preprint, not peer reviewed). https://arxiv.org/abs/2604.26539. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/narrative/027-roussilhe-2026-ict-oil-and-gas-counting-own-goals.md) | Microsoft–XTO Permian deal implies about 6.7 Mt CO₂ in 2025, 43% of Microsoft's 2024 market-based footprint; not in its Scope 3. Authors' own extrapolation from vendor claims. |
| 10 | IEA (Apr 2025). *Energy and AI*, "Energy demand from AI." https://www.iea.org/reports/energy-and-ai/energy-demand-from-ai. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/ghg/023-iea-2025-energy-and-ai-datacentre-electricity.md) | 415 TWh in 2024, about 945 TWh by 2030 (base case) |
| 11 | Rosa, L. et al. (Sep 2026). *Environmental Science & Technology*. https://doi.org/10.1021/acs.est.6c08088. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/water/036-rosa-2026-global-data-centres-water-gap-exposure.md) | 25% of 9,347 mapped data centres in cells where water demand exceeded supply at least one month a year; 33% at +3 °C |
| 12 | IEA (14 Oct 2026). *Powering Artificial Intelligence.* https://www.iea.org/reports/powering-artificial-intelligence. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/water/039-iea-2026-powering-ai-wue-measures-water-stress.md) | 34% of AI data-centre capacity in high baseline water stress; hyperscaler WUE figures use two different water measures |
| 13 | Alston, Love & Haider (21 Sep 2026). arXiv:2609.25437. https://arxiv.org/abs/2609.25437. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/water/038-alston-2026-cooling-choices-shift-water-across-grid.md) | Michigan grid model: air-cooled chillers cut a 200 MW site's total water 44%, with 96.6% of the remainder at power generators |
| 14 | Gauld, K.D., Varon, D.J., Balasus, N. & Cusworth, D.H. (Aug 2026). arXiv:2608.22153. https://arxiv.org/abs/2608.22153. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/datacenters/026-gauld-2026-satellite-nox-colossus-2-southaven.md) | Satellite-derived NOx at Southaven about 19x the permitted rate |
| 15 | Alpine, W., Alpine, H. et al. (2026). *AI-driven productivity gains enable more CO₂ emissions than they avoid in a global energy–economy model.* npj Climate Action. https://www.nature.com/articles/s44168-026-00411-0 | AI in oil and gas enables more emissions than AI avoids. **Note: the journal is npj Climate Action, part of the Nature portfolio, not *Nature*.** Axios names co-leads from Purdue University and an independent researcher. |
| 16 | Martin, A. (13 Aug 2026). "AI Applications for Oil and Gas Companies Worsen Climate Pollution." Inside Climate News. https://insideclimatenews.org/news/13082026/ai-enables-oil-and-gas-production/ | 0.47 to 1.8 Gt CO₂ a year; 1.2 to 4.8% of 2024 energy-related emissions; up to 13x the IEA's data-centre estimate; AI must lift renewables 4 to 5 times more than oil and gas to keep emissions flat |
| 17 | Axios (11 Aug 2026). "AI could help unlock more oil — and emissions." https://www.axios.com/2026/08/11/ai-oil-gas-emissions-increase | Roughly 3 to 13x IEA data-centre emissions; 64 scenarios, emissions fell only when AI gave no fossil productivity gain |
| 18 | Anthropic (28 Aug 2025). "Updates to Consumer Terms and Policy." https://www.anthropic.com/news/updates-to-our-consumer-terms | Free, Pro and Max users asked to choose whether chats train models; five-year retention for those who allow it (30 days otherwise); commercial, education, government and API excluded |
| 19 | Tom's Guide (Aug 2025). "Claude AI will start training on your data soon — here's how to opt out." https://www.tomsguide.com/ai/claude/claude-ai-will-start-training-on-your-data-soon-heres-how-to-opt-out-before-the-deadline | For existing users the large "Accept" button opted them in, and opting out took a separate toggle |
| 20 | OpenAI. "Memory FAQ." https://help.openai.com/en/articles/8590148 | ChatGPT "memory" keeps details across conversations |
| 21 | Anthropic. "API and data retention." https://platform.claude.com/docs/en/manage-claude/api-and-data-retention | Zero data retention is offered on commercial API terms |
| 22 | heise online (2026). "OpenStreetMap is concerned: thousands of AI bots are collecting data." https://www.heise.de/en/news/OpenStreetMap-is-concerned-thousands-of-AI-bots-are-collecting-data-11157359.html | About 100,000 IP addresses scraping OSM in coordination within one week, raising hosting costs |
| 23 | Boettcher, H. (29 Aug 2026). "Consumed at Scale: AI-Driven Extraction and the Political Economy of OpenStreetMap." State of the Map 2026. https://pretalx.com/sotm2026-osm-science/talk/MWCJFD/ | Dates the 100,000-IP event to late January 2026; costs fall largely on volunteers |
| 24 | Earthmover (Oct 2026). LinkedIn post, "the scientific community is rethinking the meaning and purpose of open data in the era of AI", with reply thread incl. Alex Logan (Cecil) and Olivier Mills. Text supplied by Olivier. | Frontier labs harvest open data en masse; "Simply providing unbounded anonymous access may turn out to be unsustainable"; Earthmover lets providers see users, set licences, charge for access; Logan: "does this resolve the last mile delivery issue?" |
| 25 | McKinsey & Company (25 Aug 2026). *The state of AI in 2026: On the road to ROI.* https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai | Nearly 90% use AI in at least one function; 44% scaling across the enterprise (38% a year earlier); n = 1,719 |
| 26 | Bridgespan Group with NTEN (10 Sep 2026). *Choosing Your AI Path.* https://www.bridgespan.org/press-releases/turning-ai-opportunity-into-strategy | 70% of nonprofit leaders and staff think they're missing AI opportunities; 8% have a one- to two-year AI roadmap. Sample size not stated on the press page. |
| 27 | Center for Effective Philanthropy (2025). *AI With Purpose.* https://cep.org/blog/ai-with-purpose-as-funders-and-nonprofits-adopt-ai-whats-missing | 90% of foundations offer grantees no AI adoption support; 400+ nonprofit leaders surveyed |
| 28 | Fortune (21 Oct 2026). "AI helped discover the PHP vulnerability that could take down 70% of the internet." https://fortune.com/2026/10/21/ai-tools-discovered-the-php-vulnerability-that-could-take-down-70-of-the-internet-now-the-race-is-on-to-patch-it-before-ai-powered-hackers-attack/ | Two critical PHP flaws (CVE-2026-109618, -109619) found with AI models; disclosed 21 Oct 2026; Cloudflare saw tens of thousands of exploit attempts within 45 minutes |
| 29 | Axios (22 Oct 2026). "Black October." https://www.axios.com/2026/10/22/php-vulnerability-security-update-websites | Flaws went undetected for more than a decade; PHP runs about 70% of websites whose server-side language is known; "Black October" label |
| 30 | Wikipedia. "Jev (AI model)." https://en.wikipedia.org/wiki/Jev_(AI_model) | TypeSafe AI; limited early access 15 Sep 2026; $40M seed led by DCVC; closed model |
| 31 | Latent Space / AINews (19 Sep 2026). "Here are 6 Clones of Jev in 2 days." https://latent.space/p/ainews-here-are-6-clones-of-jev-in/ | Six clones within two days (Laya, DiffusionGemmaJev, Bespoke Nimble, SemIf/OpenJev, Jevlike, Kev-0.5B); several public on GitHub/Hugging Face |
| 32 | TechCrunch (1 Oct 2026). "Amazon releases its own Jev clone as decision models flood the web." https://techcrunch.com/2026/10/01/amazon-releases-its-own-jev-clone-as-decision-models-flood-the-web/ | "Dozens of similar models"; AWS Strands Decider 2B, open source, 1 Oct 2026 |
| 33 | Wikipedia. "Shovelware." https://en.wikipedia.org/wiki/Shovelware | Software valued for quantity over quality; early-1990s CD-ROM bundles; the Wii |
| 34 | Maciejewska, J. (29 Mar 2024). Post on X/Twitter, via WIST. https://wist.info/maciejewska-joanna/81891/ | "I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do my laundry and dishes." |
| 35 | The Register (8 Oct 2024). "Eric Schmidt speech." https://www.theregister.com/2024/10/08/eric_schmidt_speech/ ; also Windows Central, same week | Schmidt: "we're not organized" to hit climate goals, "I'd rather bet on AI solving the problem than constraining it and having the problem." Used as the third "promise" line. |
| 36 | The Mary Sue (6 Oct 2026), on *The Daily Show*, "Jon Stewart Unpacks AI's Dubious Promise of a Better Future." https://www.themarysue.com/jon-stewart-calls-out-industry-giants-claiming-ai-will-save-humanity-and-his-breakdown-of-their-current-priorities-is-impossible-to-ignore/ | Quotes: "mostly murder, surveillance, and surge pricing for the McRib"; "If we let them do enough unfettered capitalism, we can apparently capitalism our way to socialism"; executives' "universal high income" promise |
| 37 | Kobak, D. et al. (2025). "Delving into LLM-assisted writing in biomedical publications through excess vocabulary." *Science Advances*. arXiv:2406.07016. https://arxiv.org/abs/2406.07016 | At least 13.5% of 2024 biomedical abstracts show signs of LLM processing, across 15.1 million PubMed abstracts (2010-2024). |
| 38 | EdArXiv preprint (posted 12 Aug 2026). https://osf.io/preprints/edarxiv/n29zd_v3, reported by AI Weekly (23 Oct 2026). https://aiweekly.co/alerts/edarxiv-audit-faults-all-14-ai-in-education-meta-analyses | Audit of 14 AI-in-education meta-analyses: none valid; 2,000+ citations combined. The report does not say AI was used in the audit. |
| 39 | *Matter of Newby v Adelphi Univ.*, 2026 NY Slip Op 26021 (Sup. Ct. Nassau County, 28 Jan 2026, Marber, J.). https://law.justia.com/cases/new-york/other-courts/2026/2026-ny-slip-op-26021.html | Turnitin gave a student's essay a 100% AI-generated score; the court annulled the violation and ordered the record expunged ("without valid basis and devoid of reason"). Note: the ruling turned on process and the professor's own emails, not on a finding about Turnitin's accuracy. |
| 40 | Curtin University (4 Sep 2025). "Update on Turnitin AI detection tool." https://www.curtin.edu.au/news/oasis-news/update-on-turnitin-ai-detection-tool/ | Curtin disables Turnitin's AI-writing detection across all campuses from 1 January 2026; text-matching stays |
| 41 | CSET, Georgetown (7 Dec 2022). "Profile highlight: Hugging Face." https://cset.georgetown.edu/publication/profile-highlight-huggingface/ | More than 80,000 models on Hugging Face in December 2022 (first bar of the model-count chart) |
| 42 | Mind-Verse (19 Aug 2026), citing Hugging Face on X (18 Aug 2026). https://www.mind-verse.de/news/hugging-face-hub-meilenstein-3-millionen-ki-modelle | Hugging Face Hub passed 3 million public models in Aug 2026; 1 million in Sep 2024 |
| 43 | The Register (15 Oct 2025). "Schleswig-Holstein completes move to open source email." https://www.theregister.com/2025/10/15/schleswig_holstein_open_source/ | 40,000+ accounts moved from Exchange/Outlook to Open-Xchange and Thunderbird |
| 44 | The Register (27 Jan 2026). "France to replace US videoconferencing wares with sovereign alternative." https://www.theregister.com/2026/01/27/france_videoconferencing_visio/ | France's state-built Visio to replace Zoom, Teams, Webex and Meet across the public sector by 2027. |
| 45 | Unsloth documentation. https://unsloth.ai/docs | Trains models "2× faster with 70% less VRAM". |
| 46 | Baobab Tech (Oct 2026). Fine-tuned tagger experiment, internal findings (unpublished). | LLM tagger about $500–720 a month at list price. A fine-tuned EmbeddingGemma-2 with LoRA (280M parameters) scored 0.648 micro-F1 against human reviewers, against 0.468 for the LLM, and put the right tag first 88% of the time against 61%. Training cost $3.26; it tags 130–700 articles a second on one GPU, against about 2 a second via API. Prices are list-price estimates, not invoices. |
| 47 | Baobab Tech (2026), internal project data (unpublished) | Evidence-tagging model for evaluation reports: built in under two days, 10,000x faster than the frontier-model approach, no cloud GPU, $40 of experiments, under $2 final training run |
| 48 | Baobab Tech. *EvalExplorer classifier fine-tuning*, README and experiments dataset. https://huggingface.co/datasets/baobabtech/evalexplorer-classify-experiments | Fine-tuned Qwen3.5-2B reproduces pipeline labels at 0.847 mean field score; $0.89 per fine-tune (21 min, A100); 4B model as a 2.8 GB GGUF at 0.841. Labels are silver (unreviewed LLM output). |
| 49 | Baobab Tech. *Water Conflict Classifier* collection, Hugging Face. https://huggingface.co/collections/baobabtech/water-conflict-classifier | Two models (33.4M and 22.7M parameters); training dataset of about 3,500 rows; evals published |
| 50 | Baobab Tech. *water-conflict-classifier* README and PyPI package. https://pypi.org/project/water-conflict-classifier/ | v2.5 (Dec 2025): 33M params (bge-small, SetFit), about 2,000 training examples, 81.9% exact match, F1 micro 0.866, Weapon F1 0.634; 10.6 ms per headline in demo. |
| 51 | Vivek, V. (7 Jul 2026). "Microsoft pushed Copilot everywhere, but barely anyone bought it, and even fewer use it." Digital Trends, citing Fortune (29 May 2026). https://www.digitaltrends.com/computing/microsoft-pushed-copilot-everywhere-but-barely-anyone-bought-it-and-even-fewer-use-it-report/ | 20 to 30% of licensed Microsoft 365 Copilot seats used weekly, from enterprise surveys cited by Fortune |
| 52 | Baobab Tech. *decision-models-experiments*, README and docs/landscape.md. https://github.com/baobab-tech/decision-models-experiments | 255 decision models catalogued in Han Xiao's *All about Jev* dataset by its 27 Sep 2026 sweep, 12 days after Jev's 15 Sep launch; Baobab Tech's own zero-shot and fine-tuning experiments on evaluation-report excerpts |
| 53 | Holzwarth, L., González-Márquez, R. & Kobak, D. (11 Aug 2026). "Most biomedical publications show signs of LLM-assisted writing." arXiv:2608.10715 (preprint, not peer-reviewed). https://arxiv.org/abs/2608.10715 | Unbiased word-frequency estimate on 1,194,287 open-access PubMed Central papers (2017-2025). Abstracts: 9% (2023), 31% (2024), 53% (2025), 68% (Dec 2025). Full papers: 19%, 52%, 77%, 89%. Supersedes the 13.5% lower bound in [37], which used a different, conservative method; the two are not directly comparable. |
| 54 | Sharma, Tiwari, Kaur, Park & Pinsky (1 Sep 2026). "The water cost of intelligence." *Green* (MDPI) 1(2):8. https://doi.org/10.3390/green1020008. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/water/027-sharma-2026-water-cost-of-intelligence-boundary.md) | Reconstructs Google's 0.26 mL per median Gemini prompt (on-site cooling only) and adds 0.471 mL of electricity-generation water for 0.725 mL total, +179%. A long Gemini prompt reaches about 28 mL. All inputs are Google's self-reported figures; nothing measured independently. |

## Further Reading

- BetaKit (13 May 2026). "Ex-Hugging Face, Salesforce leaders partner on AI sustainability venture." https://betakit.com/ex-hugging-face-salesforce-leaders-partner-on-ai-sustainability-venture/. Luccioni and Gamazaychikov founded the Sustainable AI Group (Montréal), May 2026
- Enabled Emissions Campaign. https://www.enabledemissions.com/. Will and Holly Alpine's nonprofit, founded 2024 (founding year per Axios [17])
- Meta Platforms (Oct 2025). *2025 Environmental Data Index.* https://sustainability.atmeta.com/wp-content/uploads/2025/10/Meta_2025-Environmental-Data-Index.pdf. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/washing/020-meta-2025-scope2-4394x-gap.md) 2024 Scope 2: 1,358 t market-based, 5,967,348 t location-based (4,394x)
- Google (2026). *2026 Environmental Report.* https://sustainability.google/files/google-2026-environmental-report.pdf. Research notes: [1](https://github.com/baobab-tech/research-ai-energy/blob/main/research/washing/018-google-2026-market-vs-location-based-gap.md), [2](https://github.com/baobab-tech/research-ai-energy/blob/main/research/washing/023-water-positive-pledge-accounting-2026.md), [3](https://github.com/baobab-tech/research-ai-energy/blob/main/research/narrative/022-google-2026-enabled-emissions-41mt.md), [4](https://github.com/baobab-tech/research-ai-energy/blob/main/research/water/031-google-2026-fy2025-water-disclosure.md) Operational emissions down 2.3% while electricity use rose 37%; replenishment at 78% of freshwater consumption; 41 Mt "AI solutions", 27.8 Mt (68%) from Google Earth, "not independently verified"
- Microsoft (Jul 2026). *2026 Environmental Data Fact Sheet* (FY25). https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/microsoft/msc/documents/presentations/CSR/2026-Microsoft-Environmental-Data-Fact-Sheet-PDF.pdf. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/water/030-microsoft-2026-fy25-site-level-water.md) "Water positive" overall; delivered replenishment at 7 of 28 data-centre locations
- Amazon (2026). *2025 Amazon Sustainability Report.* https://sustainability.aboutamazon.com/2025-amazon-sustainability-report.pdf. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/washing/023-water-positive-pledge-accounting-2026.md) 75% progress to water-positive goal for data centres
- IEA (Apr 2025). *Energy and AI*, "AI and climate change." https://www.iea.org/reports/energy-and-ai/ai-and-climate-change. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/narrative/021-iea-2025-energy-and-ai-avoided-emissions-scenario.md) 1,400 Mt avoided-emissions case; "could be marginal", "could be negated by rebound effects"
- Carmeno, N.L., Domingos, T. & O'Neill, D.W. (2026). arXiv:2602.24091. https://arxiv.org/abs/2602.24091. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/narrative/028-carmeno-2026-systematic-review-83-percent-positive.md) 83% of environmental studies portray AI's impact as positive; 11% consider systemic effects (rebound, induced demand, structural change).
- Monitoring Analytics (5 Jan 2026). *Analysis of the 2027/2028 RPM Base Residual Auction, Part A.* https://www.monitoringanalytics.com/reports/Reports/2026/IMM_Analysis_of_the_20272028_RPM_Base_Residual_Auction_Part_A_20260105.pdf. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/community/016-monitoring-analytics-2026-pjm-capacity-data-center.md) $23.1bn of capacity cost across three auctions attributed to data-centre load
- Wilkins, J. (8 Oct 2026). "Adobe in Trouble After Someone Reverse Engineered Adobe Creative Suite and Released It as Open Source." Futurism. https://futurism.com/future-society/adobe-creative-suite-software-open-source-artcraft. ArtCraft: four-person team says it used Claude to build free, open-source Adobe-like apps; some in early alpha. The Claude claim is the team's own account.
- Virtuous & Fundraising.AI (2026). *2026 Nonprofit AI Adoption Report*, via NonProfit PRO (17 Feb 2026). https://www.nonprofitpro.com/article/nonprofit-ai-adoption-hits-92-but-only-7-see-major-impact. Backup only: 92% use AI, 7% see major impact, 4% have documented workflows; vendor survey of 346 nonprofits
- Atherton, D. (11 Aug 2026). "Incident Report 2026: May, June, July." AI Incident Database (Responsible AI Collaborative). https://incidentdatabase.ai/blog/incident-report-2026-may-june-july/. 148 new incident IDs added May to July 2026; largest groups scams/impersonation and political influence. **Caveat:** IDs added in the period, not incidents that occurred in it.
- The White House (Sep 2026). "Fact Sheet: President Donald J. Trump Inaugurates The Era of Super Intelligence." https://www.whitehouse.gov/fact-sheets/2026/09/fact-sheet-president-donald-j-trump-inaugurates-the-era-of-super-intelligence/. "Super intelligence" executive order
- Nextgov/FCW (29 Sep 2026). "White House unveils 'super intelligence' executive order and industry accord." https://www.nextgov.com/artificial-intelligence/2026/09/white-house-unveils-super-intelligence-executive-order-and-industry-accord/416325/. EO and voluntary frontier safety accord built on the labs' own internal controls and audits; administration opposed a dedicated regulator.
- Dahl, J. (13 May 2026). "Is Growth Always Good News? 2026 Article Submission Surges." The Scholarly Kitchen. https://scholarlykitchen.sspnet.org/2026/05/13/guest-post-is-growth-always-good-news-2026-article-submission-surges/. ScholarOne journals: submissions up 33% in Q1 2026 year on year; desk rejections up 72% from 2022 to 2025; "The 2026 submission surge is a stress test."
- heise online (Mar 2025). "No more calculation errors? AI checks papers for errors." https://heise.de/-10321469. Black Spatula Project: LLMs check papers for errors, about 500 checked, about 10% false alarms; YesNoError analysed 37,000+ papers
- Pacific Institute. *Water Conflict Chronology* fact sheet. https://pacinst.org/wp-content/uploads/2024/08/Water-Conflict-Chronology_Fact-Sheet.pdf. The chronology the classifier supports
- Open Source For You (Feb 2026). "Trump tensions push EU toward open source and away from US tech." https://www.opensourceforu.com/2026/02/trump-tensions-push-eu-toward-open-source-and-away-from-us-tech/. EU tech sovereignty package covering cloud, AI and chips; EU relies on foreign suppliers for 80%+ of digital products and services
- Holland & Knight (18 Aug 2026). "Operation AI Comply: 2 Years Later." https://www.hklaw.com/en/insights/publications/2026/08/operation-ai-comply-2-years-later-continued-enforcement. FTC brought more than a dozen AI-washing cases in 2025; securities class actions filed in 2026
- Sensor Tower data via *The New York Times* (20 Jul 2026), as reported by Pasquale Pillitteri. https://pasqualepillitteri.it/en/news/8867/app-store-vibe-coding-surge. About 560,000 new App Store apps in H1 2026, nearly double the year-earlier pace; credited to vibe coding
- Press Gazette (updated 23 Oct 2026). "Journalism job cuts in 2026." https://pressgazette.co.uk/news/journalism-job-cuts-in-2026-updates/. BBC 1,800 to 2,000 jobs (April); Washington Post 300+ newsroom journalists (February); 3,434+ cuts in 2025
- Lighthouse Reports et al. (30 Sep 2026). "Data centre silence." https://www.lighthousereports.com/investigation/data-centre-silence/. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/policy/029-lighthouse-2026-eu-data-centre-reporting-foi-gaps.md) Ten of 27 EU member states hold no data-centre figures; Dutch registry lists 104 of at least 186 reportable facilities
- Raychaudhuri, D. & Volcovici, V. (14 Jul 2026). Reuters. https://www.reuters.com/world/americas/pollution-musks-unpermitted-xai-power-project-hits-hardest-black-communities-2026-07-14/. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/community/029-reuters-2026-xai-turbines-unpermitted-analysis.md) 59 unpermitted xAI turbines for Colossus 2, counted from regulator emails
- The Texas Tribune (14 Oct 2026), reporting Texas Water Development Board records. https://www.texastribune.org/2026/10/14/texas-data-centers-water-use-penalties-enforcement/. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/water/041-twdb-2026-texas-data-centre-water-survey-compliance.md) 233 of 334 data centres identified for 2020-2025 never returned a water survey; penalty a $500 misdemeanour
- Masanet, E. (3 Oct 2026). arXiv:2610.04735. https://arxiv.org/abs/2610.04735. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/water/037-masanet-2026-estimating-data-center-water-use-best-practices.md) Estimates span 24-fold for one US 2030 scenario depending on inputs; 85% of studies with indirect estimates don't state hydropower accounting
- Privette, A.P., Barros, A. & Cai, X. (27 Feb 2026). *AGU Advances*. https://doi.org/10.1029/2025AV002140. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/water/046-privette-2026-data-centers-water-footprint-transparency.md) Data centres at 21% of city water use in Council Bluffs and 29% in The Dalles; small nationally, large locally, largely undisclosed
- NCSL (8 Oct 2026). "As they consider new data centers, states target water use." https://www.ncsl.org/resources/details/as-they-consider-new-data-centers-states-target-water-use. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/policy/027-ncsl-california-2026-state-data-centre-water-laws.md) About 100 bills in 39 states on data-centre water in 2025-26; California signed two water laws, 21 Sep 2026
- European Commission (21 Sep 2026). *Report on the energy efficiency of data centres in the EU*, COM(2026) 500. https://energy.ec.europa.eu/publications/report-energy-efficiency-data-centres-eu_en. [Research note](https://github.com/baobab-tech/research-ai-energy/blob/main/research/policy/028-ec-2026-com-500-data-centre-energy-efficiency-report.md) Average reported EU WUE 0.58 m³/MWh across 458 facilities (not yet used in outline)
- Liang, W., Yuksekgonul, M., Mao, Y., Wu, E. & Zou, J. (2023). "GPT detectors are biased against non-native English writers." *Patterns*. arXiv:2304.02819. https://arxiv.org/abs/2304.02819. Seven detectors flagged non-native English (TOEFL) essays as AI-written at an average false-positive rate of 61.22% (91 essays).