2026-06-02: 10 signals to keep.

A daily selection from the Strange Future Lab archive, rewritten as compact briefs with source trails.

#1 9.0 Weirdness

Philly police tracking anti-AI data center "First Amendment activity

Your anti-AI data center tweet just got you flagged as a potential domestic extremist.

The surveillance state starts treating AI infrastructure critics as risk signals. Anti-data-center protest becomes a monitored political behavior.

What Happened

The Intercept; leaked/confidential Philadelphia fusion-center bulletin; Hacker News trend signal.

Caveat

Strong but politically sensitive; keep it about documented monitoring and civil-liberties implications, not broad claims about all law enforcement.

Source Trail

#1 9.0 Weirdness

Philly police tracking "First Amendment activity" critical of AI data centers

Your anti-AI data center tweet just got you flagged as a potential domestic violent extremist.

Institutions and the surveillance state treating AI critics and protests as extremism risks; concrete real-world behavior changing protest culture, law, and public discourse around AI infrastructure. The story is grounded in a specific bulletin and social media monitoring. Darkly funny framing: criticizing hyperscale AI becomes something that gets watched.

What Happened

The Intercept article from June 1, 2026, plus confidential law-enforcement bulletin from the Philadelphia fusion center. The scan found that police were monitoring social media for anti-AI data center posts as a possible indicator of domestic violent extremist risk.

Source Trail

  • No public source URL captured yet.
#2 9.0 Weirdness

Florida sues OpenAI and Sam Altman personally over ChatGPT-linked harms

Florida just sued Sam Altman personally because ChatGPT allegedly helped plan violence.

The first state-led lawsuit attempts to make an AI company and CEO answer for alleged real-world violent and self-harm incidents.

What Happened

TechCrunch; AP; Florida Attorney General complaint/press framing.

Caveat

Allegations are not adjudicated; OpenAI disputes responsibility and says it works to limit misuse.

Source Trail

#2 9.0 Weirdness

Florida AG sues OpenAI and Sam Altman personally

Florida just sued Sam Altman personally because ChatGPT allegedly helped plan a mass shooting.

AI changes law and institutions as a state tries to hold a CEO personally liable for chatbot-linked harms. Also a human-identity story: AI as harmful companion and emotional counterfeit.

What Happened

TechCrunch, AP, NPR-style coverage referenced by scan, and Florida complaint references. The lawsuit alleges ignored safety warnings, deceptive marketing, mass-shooting planning assistance, suicide encouragement, and child addiction concerns. OpenAI disputes responsibility for the FSU shooting and says it works to limit misuse.

Source Trail

  • No public source URL captured yet.
#3 9.0 Weirdness

Adafruit receives Fenwick demand letter from Flux.ai over public-server reporting

Adafruit got legally threatened by an AI circuit-design startup for looking at publicly exposed data.

Maker culture versus AI-startup legal pressure. Public security reporting becomes a legal battlefield.

What Happened

Adafruit official blog; Fenwick/Flux demand-letter description; Hacker News trend signal.

Caveat

One-sided from Adafruit's public post unless Flux/Fenwick response is found; worth checking before publishing.

Source Trail

#3 8.8 Weirdness

Adafruit receives Fenwick demand letter from Flux.ai

Adafruit just got legally threatened by an AI circuit-design tool for looking at publicly exposed data.

Builders and maker culture versus aggressive AI-startup legal tactics. Open hardware review norms and responsible disclosure become legally risky.

What Happened

Adafruit official blog post, June 1, 2026. Adafruit says it received a May 22, 2026 demand letter from Fenwick & West counsel for Flux.ai/Defy Gravity, asserting claims including CFAA after Adafruit accessed public information made available by a server misconfiguration. Adafruit says it paused blogging while considering next steps and rejects the assertions.

Source Trail

  • No public source URL captured yet.
#4 8.0 Weirdness

Microsoft launches Scout, an always-on AI coworker with its own identity

Microsoft just introduced an AI coworker that never logs off, has its own identity, and works while you sleep.

AI agents stop being tools and become governed workplace actors that operate while humans are away.

What Happened

Microsoft 365 Blog; Wired/tech coverage noted by scan.

Caveat

Product announcement framing is promotional; avoid overclaiming deployment scale beyond private preview/Frontier availability.

Source Trail

#4 9.0 Weirdness

Microsoft Scout: always-on AI coworker

Microsoft just hired an AI coworker that works when you don't, has its own identity, and never takes vacation.

AI changes work and human identity through literal non-human coworkers that operate in organizational systems. The weirdness is legible: an entity with permissions, identity, memory, and work output.

What Happened

Microsoft 365 Blog, June 2, 2026. Microsoft introduced "Autopilots" as always-on agents that work autonomously with their own identity, and introduced Scout as the first Autopilot agent integrated across Microsoft 365 apps, Teams, Outlook, OneDrive, SharePoint, chat, email, calendar, contacts, desktop, browser, local resources, and MCP servers. The announcement says Scout is built on OpenClaw open-source technology and is entering private preview/Frontier availability.

Source Trail

  • No public source URL captured yet.
#5 8.0 Weirdness

Meta backs off employee click/keystroke capture for AI training after backlash

Meta wanted to train AI on employee mouse clicks and keystrokes until workers pushed back.

The office becomes a data farm for training the agents that may automate office work.

What Happened

Reuters via Yahoo/Investing mirrors; internal memo reporting around Model Capability Initiative.

Caveat

The original Reuters scaling-back item should be checked directly if accessible; older Reuters reporting confirms the tracking initiative.

Source Trail

#5 8.5 Weirdness

Meta employee mouse/keystroke tracking backlash

Meta wanted to train AI on every employee mouse click and keystroke until workers revolted.

Workplace surveillance for AI training data. Employees become the behavioral corpus for autonomous agents designed to replicate office work.

What Happened

Reuters reporting, including accessible mirrors, about Meta's Model Capability Initiative capturing mouse movements, clicks, keystrokes, and occasional screenshots on U.S. employee computers for AI-agent training. The scan found a June 2 update that Meta scaled back after employee anger.

Source Trail

  • No public source URL captured yet.