How Much Water Does an AI Prompt Use? Why Estimates Differ
AI water-use figures depend on the model, task, location and accounting boundary. Learn how to compare estimates without a misleading universal number.
Research archive
Technology & privacy
Evidence on AI, privacy, digital platforms and the companies building them.
39 articles
Page 1 of 2
Phone Energy Labels Explained: Battery, Drops and Repairability
Learn to read the EU smartphone label and EPREL record. Energy efficiency, battery endurance and repairability answer different buying questions.
Replace the Battery or Buy a New Phone? A Decision Checklist
Compare a phone battery replacement with a new purchase using repair cost, remaining software support, other faults and the phone’s usefulness to you.
How Ethical Is the iPhone Duo? What Launch Evidence Shows
Apple’s folding iPhone has documented recycled-material claims, but repairability and factory conditions still need scrutiny. Here is what preorder buyers can verify.
Phone Software Support: Count Years Left, Not Years Advertised
Seven years of updates does not mean seven years from your purchase. Separate security support, OS upgrades and repair availability before buying.
Buying a Refurbished Phone: Seven Checks Beyond the Grade
Check support years, battery condition, locks, repair history and seller terms before buying a refurbished phone. Cosmetic grades are only one clue.
Do ChatGPT and Claude Train on Your Chats? Check These Settings
Compare ChatGPT and Claude training controls. Paid personal plans, chat history, retention and business workspaces have different privacy rules.
Fairphone or a Refurbished Phone: Which Choice Makes Sense?
Compare a new Fairphone with a refurbished phone on repair, remaining support and useful life, without pretending one option always wins.
Microsoft, Gaza and Palestinian Surveillance: What the Documents Show
A sourced timeline of Microsoft’s reviews, the suspension of specific Israeli military services, and the questions its 2026 follow-up leaves open.
Google vs Microsoft: Which Company Is More Ethical?
Google and Microsoft dominate search, cloud and AI. We compare their records on privacy, workers, military contracts, climate and community impact.
AI Agents Hacked Real Companies: What the Labs Disclosed
OpenAI and Anthropic disclosed that experimental AI agents escaped test boundaries and reached real systems. Here is what happened—and what it does not prove.
AI Companies Ranked by Ethics (2026): Scores and Evidence
Which AI company is most ethical? Compare nine companies on technology, climate and business conduct, with dated scores, primary sources and clear limits.
The Most Ethical Phone Brands in 2026, Ranked by Evidence
Every smartphone carries the same supply-chain baggage. We rank Apple, Samsung, Xiaomi and Google on verified conduct across labour, sourcing, environment and privacy.
Verizon vs AT&T vs T-Mobile: Which Carrier Is Most Ethical? (2026)
Compare Verizon, AT&T and T-Mobile on privacy, trust, workers and climate. See dated ethics scores, FCC cases and practical privacy checks before switching.
Is Apple Ethical? The Full Evidence, Scored
Apple scores well on environment and privacy but poorly on labour and business conduct. We score the company across 11 values, with the evidence behind each.
NVIDIA vs AMD: Which Chipmaker Is More Ethical? (2026)
Two AI-chip giants, two ethical records. We compare NVIDIA and AMD across 11 values to see which is the more defensible holding after the June 2026 selloff.
Is Amazon Ethical? What the Evidence Says
Amazon scores positively on health and renewable energy but fails on labor, sourcing, and business conduct. The evidence from court filings and regulatory actions points both ways. Here is the full record.
Is Salesforce Ethical? +40 on Honesty — Here's What That Means
Salesforce has one of the highest Honest & Fair Business scores in our database. But it also scores -50 on weapons. The full picture.
The 2026 IPO Wave: Ethics Scores for Every Company Going Public
SpaceX, OpenAI and Anthropic are racing to public markets at a combined valuation near $4 trillion. Here is what court filings and regulatory records say about each of them before you buy.
Anthropic Is Going Public. Does the Safety-First AI Lab Actually Score Safer?
Anthropic filed its confidential S-1 on June 1 at a $965 billion valuation. We checked whether its safety-first branding survives contact with independent evidence. Mostly, it does.
Is OpenAI Ethical? The $1 Trillion IPO, the Florida Lawsuit, and the Evidence.
OpenAI confirmed a $1 trillion IPO seven days after Florida sued it over ChatGPT safety. What the court filings say about the company's conduct.
xAI's Regulatory Rap Sheet: Grok, Deepfakes and Two Government Investigations
Grok generated an estimated 1.8 million sexualized images of women in nine days. Two regulators opened formal investigations, 35 attorneys general demanded action, and the lawsuits keep coming. Here is xAI's record, scored.
Intel vs AMD: Which Is More Ethical? (2026 Chip Comparison)
Intel and AMD make competing chips. But Intel scores +40 on Fair Pay and Fair Trade while AMD hovers near zero on almost everything.
Is Palantir Ethical? -80 on Weapons, +30 on Health
The same AI company that powers military targeting also helps hospitals fight disease. A 110-point gap between its best and worst scores.
The Evidence Brief