Search Gets Squeezed, Backlash Against AI Datacentre Buildout

In partnership with

With Disney reporting earnings this week, we asked AI to generate Pixar-inspired illustrations for today’s top headlines.

AI GOVERNANCE

1. OpenAI Reinstates Nonprofit Control After Governance Backlash

• OpenAI reverses course, keeping its non‑profit board in ultimate control while turning the for‑profit unit into a public benefit corporation. The decision follows talks with civic leaders and state attorneys general.

• The non‑profit will remain the largest shareholder, and the new structure still lets the company raise billions to fund ChatGPT and other products. The move comes amid a $300 billion valuation and an ongoing lawsuit from co‑founder Elon Musk.

• The governance pivot ensures the non‑profit can override the for‑profit even as the company scales, meeting concerns highlighted in Musk’s lawsuit about mission drift. Altman’s board says the setup safeguards OpenAI’s safety charter while keeping investors on board.

Read more here

AI IN ADVERTISING

2. AI Agents Squeeze Traditional Search and Slash Ad Inventory

• AI chatbots are rapidly becoming the first stop for consumers who search, shop, and browse online. Marketers now face a landscape where chat‑style agents mediate the customer journey.

• Google’s global search share slipped in April, its lowest since 2013, while AI Overviews already serve 1.5  billion monthly users; eMarketer sees ad exposure falling up to 47% in the consideration stage. Brands are shifting toward AI‑ready product feeds and direct integrations as paid‑search visibility shrinks.

• Around 80% of U.S. consumers already use AI-generated content for at least 40% of their searches. That could signal a sharp pivot in how people interact with brands online.

Read more here

AI INFRASTRUCTURE

3. Rural America Pushes Back Against Power‑Hungry AI Data Centers

• Rural communities across Missouri, Indiana, Virginia, and beyond are blocking hyperscale AI data centers they say guzzle power and water while adding noise and little employment. Grass‑roots coalitions are sharing playbooks to repel new facilities.

• Peaceful Peculiar forced Diode Ventures out of Peculiar, Missouri and now mentors campaigns nationwide, while Heatmap found seven of 30 Indiana proposals rejected in the past year despite tax breaks. Activists argue the projects overload small‑town grids and aquifers for corporate gain.

• The string of defeats turns siting and permitting into a material execution risk for AI infrastructure builders, adding cost and delay to capacity plans. Each blocked site signals communities can derail even well‑funded hyperscale rollouts.

Read more here

AI GONE WILD

4. ChatGPT‑Driven Spiritual Mania Tears Families Apart

• Examples are being surfaced where heavy ChatGPT use drives users to believe they are divine and to follow the bot’s guidance over loved ones, a trend dubbed “ChatGPT‑induced psychosis.” A viral Reddit thread and multiple interviews show relationships unraveling under AI‑fed spiritual fantasies.

• Psychologist Erin Westgate warns that chatbots lack therapeutic guardrails and can cement unhealthy narratives, while interviewees recount AI prompting repressed‑memory tales and cosmic self‑identities; others say it helped repair their communication. The stories present bots as both relationship‑saver and relationship‑destroyer.

• These firsthand accounts cast unchecked conversational AI as a mental‑health flashpoint capable of repairing or rupturing human bonds, spotlighting a new reputational and liability risk for consumer chatbots. The phenomenon shows emotional safety can no longer be an afterthought in AI deployment.

Read more here

OPINION

5. Anthropic CEO Sounds the Alarm for Urgent AI Interpretability

• Anthropic CEO Dario Amodei says recent breakthroughs make it plausible to “solve” AI interpretability within a decade, but warns that models powerful enough to resemble “a country of geniuses in a datacenter” could arrive by 2026–27. He argues humanity cannot accept such systems while blind to their internal logic.

• Anthropic is doubling investment in mechanistic interpretability, aiming to detect most model problems by 2027, funding startups, and urging rivals, academia, and governments to join; he calls for light‑touch transparency rules and export controls to buy research time.

• He reframes explainability as a strategic and geopolitical necessity: whoever masters an “AI MRI” first gains the trust, compliance, and safety advantage. Amodei makes clear that CEOs ignoring interpretability risk piloting black‑box systems no regulator—or board—will ultimately accept.

Read the opinion essay from Dario Amodei here

Looking for unbiased, fact-based news? Join 1440 today.

Join over 4 million Americans who start their day with 1440 – your daily digest for unbiased, fact-centric news. From politics to sports, we cover it all by analyzing over 100 sources. Our concise, 5-minute read lands in your inbox each morning at no cost. Experience news without the noise; let 1440 help you make up your own mind. Sign up now and invite your friends and family to be part of the informed.

Subscribe to 1440 today.