No Water? Drink Data!

In partnership with

In the Indian city of Visakhapatnam this month, activists and children marched against the $15 billion Google-Adani AI campus rising on the edge of their city, carrying banners that read "We cannot drink DATA" and painting handcuffs onto the Google logo.

On Monday 24 August, they lost the first round in court. A bench headed by Chief Justice Lisa Gill refused the interim status quo that would have frozen work at the site, and instead told the Andhra Pradesh government to ensure construction proceeds only once every permission is in hand. Every department that signed off must file a consolidated response, and the matter returns to court next week.

The state said it had followed the law throughout. The court was less relaxed, saying environmental clearance cannot be given a go-by, and pressing the government on the water and electricity the 1 GW facility will draw and on the roughly ₹22,000 crore (USD 2.30 billion) in incentives attached to it.

The PIL was brought by Bolisetty Satyanarayana. Around 160 acres of the land belongs to Simhachalam Devasthanam, a temple trust, and he argues temple land cannot be alienated without the High Court's permission. The state says nothing was transferred, only leased to the Andhra Pradesh Industrial Infrastructure Corporation, in a deal worth about ₹26 crore (USD 2.72 million) a year to the trust. Reporting on a February 2026 state order puts the full allotment far higher, at roughly 601 acres across three sites, a figure that appears only in trade and regional coverage, not the court record.

The petition also places the site 0.86 km from Kambalakonda Wildlife Sanctuary, close enough to demand stricter scrutiny, and says the existing clearance covers putting the building up rather than running a data centre once it is built.

Neil Carter, who studies sensory pollution at the University of Michigan, told National Wildlife that AI data centres could become "sensory danger zones" where light and noise pass the point at which animals stop functioning normally. His own work has tracked how human-made sound suppresses breeding in birds like the northern cardinal and the barn swallow.

Handled a lot differently in the US

On 18 August, Pennsylvania Governor Josh Shapiro signed Executive Order 2026-05, barring the state's environmental regulator from issuing a data centre permit until the developer has made a legally binding commitment to Pennsylvania's GRID standards and secured local approval. AI data centres were also stripped out of fast-track permitting, and non-disclosure agreements on these projects banned.

The same day in Nevada, the Nye County Board of Commissioners voted 5-0 to ban data centres outright in Hydrographic Basin 162, one of the most over-appropriated water basins in the state, killing a proposed 900-acre campus near Pahrump. It is the first municipal ban in Nevada.

In the US, local consent is now a gate the developer has to pass. In India, local consent is yet to legally matter in data centre constructions.

“Pervert Glasses” Taken To Court

On 24 July, a man sitting alone in a café in Delhi's Khan Market was accosted and insulted by a stranger who followed him across the room until he moved seats. He did not know he was being filmed. The stranger was a content creator wearing Ray-Ban Meta smart glasses. The reel went up on Instagram on 26 July and drew over 419,000 views and hundreds of mocking comments.

Read: Meta’s Smart Glasses Face Their First Legal Challenge In India Over Covert Recording

He tried to report it, but found no category in Instagram's menu for a privacy violation or a non-consensual recording. He chose bullying and harassment as the nearest fit, but the platform twice told him the video did not breach its Community Standards, then rejected his appeal in identical wording. Meta's grievance officer for India never acknowledged his case. The account vanished around 9 August and was back by 16 August with the video still up.

On 24 August he sent a legal notice to Meta and the creator, demanding a takedown, an apology and ₹2.05 crore (USD 214,660) in damages, with advocate Apar Gupta and the Internet Freedom Foundation acting for him pro bono.

Meta's usual shield in India is Section 79 of the IT Act, which protects intermediaries from liability for what users post. The notice walks around it by treating Meta as a manufacturer, arguing the claim arises from its design and marketing choices about a physical product and sits outside safe harbour altogether. If a court accepts that, Meta can be sued for building the glasses rather than for hosting the video. No Indian court has ruled on it yet. Meta told Scroll the glasses ship with privacy features including the recording light, and that it removes harassing content whatever device filmed it.

Germany wants them off the shelves

On 12 August, the digital rights group HateAid filed a criminal complaint with the federal Central Office for Combating Internet Crime against Meta, Ray-Ban, Oakley and several eyewear retailers, seeking a ban on sales of the Ray-Ban Meta Wayfarer. HateAid says the glasses are indistinguishable from ordinary eyewear in a country with 41 million spectacle wearers, disguising surveillance technology as an everyday object in breach of German telecommunications and data protection law. Failing a ban, it wants colour-coded cameras.

Meta's answer to all of this is a small LED that blinks while the glasses record, plus tamper detection added in July that is supposed to cut the camera if that light is covered. CBS News correspondent Anna Schecter put it to the test. She started recording, then covered the light, but the camera carried on. Meta said the feature is designed to catch an obstruction before the wearer presses record.

Israel Games LLMs With Fake Thinktank

The Guardian reported on 26 August that a website set up and funded by the Israeli government, publishing under the name of the Hanover Institute for Public Policy, put out 124 reports and more than 560,000 words between 6 and 14 August.

The institute appears not to exist as a legal entity in any jurisdiction. It has no address, no named staff and no bylines, but it does leave a paper trail.

The New York media production company Piro Inc registered the material with the US Justice Department under the Foreign Agents Registration Act as content it distributes for the Israeli government, part of a wider multimillion-dollar Israeli campaign to shape online discourse on Israel and Gaza.

The reports are dressed as neutral policy research, citing the World Bank, UN agencies and Amnesty International, on genocide allegations, war crimes, the treatment of Palestinian prisoners and starvation in Gaza.

In 20 reports, human rights findings on genocide or apartheid are answered by noting that no court has entered a final verdict. In 55, casualty and aid figures are attributed to "a party to the events".

Most of the headlines are phrased as questions, shaped like something a person would type into a chatbot, and the site runs on a commercial platform that sells content optimisation for AI citation. Nick Cleveland-Stout of the Quincy Institute, who investigated it, told the Guardian the site looks academic until you read it closely, and that seeding material for language models to absorb is now within reach of any government or company.

MESSAGE FROM OUR SPONSOR

Cut Lead Review From Hours To Minutes

Sign up for a free trial of Attio, the agentic CRM.

Ask Attio to build a daily workflow that surfaces the deals that need your attention today, like anything with a stage change, a recent reply, or a new signal in the last 24 hours.

Review your pipeline in Claude, synced live from Attio via MCP.

That's it.

Try Attio Now

📬 READER FEEDBACK

💬 What are your thoughts on using AI chatbots for therapy? If you have any such experience we would love to hear from you.

Share your thoughts 👉 [email protected]

Have you been targeted using AI?

Have you been scammed by AI-generated videos or audio clips? Did you spot AI-generated nudes of yourself on the internet?

Decode is trying to document cases of abuse of AI, and would like to hear from you. If you are willing to share your experience, do reach out to us at [email protected]. Your privacy is important to us, and we shall preserve your anonymity.

About Decode and Deepfake Watch

Deepfake Watch is an initiative by Decode, dedicated to keeping you abreast of the latest developments in AI and its potential for misuse. Our goal is to foster an informed community capable of challenging digital deceptions and advocating for a transparent digital environment.

We invite you to join the conversation, share your experiences, and contribute to the collective effort to maintain the integrity of our digital landscape. Together, we can build a future where technology amplifies truth, not obscures it.

For inquiries, feedback, or contributions, reach out to us at [email protected].