Sector: LLM and publishing (copyright theft)

Issue: Copyright

Year begun: 2024 (13 months as of Feb 2026)

Grievance: Copyright infringement

Time: 1 year. The legal case was initially launched by authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson in August 2024, with a settlement reached in September 2025.

Frequency: 1 legal claim

Location: US

Agent type: Artists (authors/copyright holders)

Type of activity: Lawsuit

Motivation for anti-AI: Compensation for copyright infringement

Outcome: Partial success. A settlement resulted in compensation for members of the class action, although the ruling also allowed fair use of legally obtained books. although note that the ruling also allows ‘fair use’ of books that are legally obtained (the problem was that they weren’t legally obtained – so only partial success).

• The Bartz v. Anthropic Settlement: Understanding America’s Largest Copyright Settlement

Sector: Music

Issue: Copyright

Grievance: Copyright for music

Time: 2024 to Feb 2026 (ongoing, around 18 months)

Frequency: Ongoing campaign

Location: UK

Agent type: Artists, especially musicians

Type of activity: Publicity campaign, including a silent record release by Paul McCartney, lobbying, and campaigning around the government consultation

Motivation for anti-AI: Copyright infringement

Outcome: Successful. Government backed down and dropped the proposed position on copyright – this followed an earlier announcement that the Government would decide by 18 March 2026, when it already looked likely to postpone the decision due to opposition (only 3% of consultation respondents favoured Government proposal, and House of Lords (Communications and Digital committee) also set out opposition).

Government summary here on the consultation seeing 3% of respondents favouring the Government’s preferred option

• Source 1
• Source 2
• Source 3
• Source 4
• Source 5
• Source 6

Sector: Media

Issue: Copyright

Grievance: Copyright – unauthorised use of data for AI model training

Time: August 2025 (7 months to Feb 2026)

Frequency: Not applicable

Location: Japan

Agent type: Corporation (newspaper)

Type of activity: Lawsuit

Motivation for anti-AI: Protecting copyright

Outcome: Incomplete. No ruling yet.

• Japan Times coverage
• Law.asia analysis

Sector: Education

Issue: Against the uncritical use of AI in higher education

Grievance: Deskilling, degradation of higher education, pro-industry bias, psychological harms for students, and reduced critical thinking

Time: Open petition since June 2025, with one visible action (1 month)

Frequency: 1 petition

Location: Netherlands

Agent type: Academics

Type of activity: Petition / open letter / e-petition

Motivation for anti-AI: Commercialisation of knowledge and higher education, and the undermining of critical thought

Outcome: limited impact: petition is popular (c. 1600 signatures) but no obvious indication of impact on AI use in universities – some limited coverage in the press – e.g. here

• Open Letter: Stop the Uncritical Adoption of AI Technologies in Academia
• International Resistance To AI In Education Marked In Open Letter

Sector: Education

Issue: AI in schools

Grievance: Negative impact on learning, cognitive development, and data privacy

Time: November 2025 to Feb 2026

Frequency: Ongoing campaign

Location: USA

Agent type: Family members / parents

Type of activity: civic campaign (open letter and attended School committee), – open letter here

Motivation for anti-AI: Creating a rulebook for ethical AI in education

Outcome: Limited impact. The aim is a ban and digital literacy curriculum, but although the School Committee acknowledged concerns there has been no substantive change.

• Source 1
• Source 2
• Source 3
• Source 4
• Source 5
• Source 6

Sector: General

Issue: AI safety and the call for red lines to safeguard AI

Grievance: Risk to species

Time: Launched in November 2025 at the UN General Assembly, with 3 months of activity to February 2026

Frequency: Ongoing, but one act

Location: Global (online)

Agent type: Celebrities

Type of activity: Online statement calling for red lines

Motivation for anti-AI: Risk to species

Outcome: No visible impact so far beyond limited media attention – – there was hope that a commitment to red lines could take place at the AI summit in Feb 2026 , but no obvious results reported.

• We urgently call for international red lines to prevent unacceptable AI risks.
• The Global Call for AI Red Lines – YouTube
• Nobel Prize winners call for binding international ‘red lines’ on AI
• Red lines for AI (BBC)

Sector: Politics / regulation

Issue: AI safety (risks of Super Intelligent AI)

Grievance: Risk to species and risk of extinction

Time: 12 months to Feb 2026

Frequency: consistent lobbying – appears to begin in Feb 2025

Location: UK

Agent type: Campaign group / non-profit organisation

Type of activity: Publicity campaign, online statement, lobbying MPs, and petition

Motivation for anti-AI: Fear of extinction

Outcome: Partial success. Over 100 supporters including many MPs, but there is no sign yet of the binding regulation on superintelligent AI that the statement calls for.

• Source 1
• Source 2

Sector: Chatbots

Issue: AI safety

Grievance: unsafe product design – AI as cause of suicide

Time: 2024 to 2026, filed in November 2024 and settled in 2026 (around 14 months)

Frequency: Ongoing

Location: US

Agent type: Family member, the mother

Type of activity: Lawsuit

Motivation for anti-AI: unsafe product design – alleges chatbot caused suicide

Outcome:largely successful – caused major international press coverage, significantly raised awareness, Senate hearing, court denied defendants motion to dismiss, and dismissed First Amendment argument (meaning chatbot is a product under law and co-founders of character.AI are defendants) – all these details are covered here: https://www.humanetech.com/case-study/litigation-case-study-character-ai-and-google  – in October 2025 character-ai  announced it would restrict chat for under-18s – in January 2026 character.ai settled  (as part of a wider group of cases that were settled at same time – reiterating commitment to safety) although by this point it seems to have become a class action lawsuit?). Although as it didn’t reach a final verdict we cannot say that the claim was upheld by courts and still the question of whether chatbots are products or ‘service delivery’ – which carry different legal liabilities – and the question of free speech – all haven’t been determined yet in a final court ruling, so this piece argues we should expect plenty more lawsuits to come – hence largely successful.

• Source 1
• Source 2
• Source 3
• Source 4
• Source 5
• Source 6
• Source 7
• Source 8
• Source 9
• Source 10
• Source 11

Sector: Chatbots

Issue: AI safety

Grievance: unsafe product design – AI as cause of suicide

Time: 2025, filed in August 2025 (around 7 months to Feb 2026)

Frequency: Ongoing

Location: US

Agent type: Family members

Type of activity: Lawsuit

Motivation for anti-AI: Unsafe product design and claims that the chatbot caused suicide

Outcome: Partial success. The case has not been heard yet (as of Apr 2026) – but already OpenAI have announced changes to safety. Case discussed at Senate hearing. Caused major international press coverage, significantly raised awareness, Senate hearing – all these details are covered here: https://www.humanetech.com/case-study/litigation-case-study-openai – the FTC launched an investigation in part as a result of the Raine v. openAI case.

• Source 1
• Source 2
• Source 3
• Source 4

Sector: Chatbot

Issue: AI safety

Grievance: unsafe product design – murder and suicide resulting from AI caused psychosis; claim is psychosis caused by AI chatbot

Time: lawsuit filed in November 2025 (4 months to Feb 2026)

Frequency: Not applicable

Location: USA

Agent type: Family members, the estate of the deceased person

Type of activity: Lawsuit

Motivation for anti-AI: unsafe product design – alleges caused psychosis, murder, suicide

Outcome: Incomplete. The hearing is still to take place.

• Source 1
• Source 2
• Source 3
• Source 4
• Source 5

Sector: Chatbots

Issue: AI safety – ChatGPT suicides

Grievance: Unsafe product design, alleging ChatGPT caused suicides

Time: 2025 (4 months to Feb 2026)

Frequency: Seven lawsuits

Location: US

Agent type: Family members, brought by Social Media Victims Law Center and Tech Justice Law Project

Type of activity: Lawsuits

Motivation for anti-AI: Unsafe product design

Outcome: Incomplete. The cases had not yet been heard as of Apr 2026.

• Associated Press coverage
• New York Times report
• Wall Street Journal coverage
• Social Media Victims Law Center statement

Sector: General tech

Issue: AI safety (Negative effects of AI on society)

Grievance: risks of AI (Multiple: jobs, child mental health, ecology, discrimination, monopoly)

Time: 2026 —

Frequency: 1 (as of Feb 2026)

Location: UK

Agent type: Grassroots organisers

Type of activity: March / demonstration (took place end-Feb 2026)

Motivation for anti-AI: Establishment of a citizen’s assembly to govern AI in the UK

Outcome: Incomplete.

• https://pulltheplug.uk
• I checked out one of the biggest anti-AI protests yet

Sector: Big Tech

Issue: Military and surveillance use of Microsoft services by the Israel ministry of defence, especially Azure data storage reportedly used to power Israeli AI technology to surveil and target Palestinians

Grievance: Opposition to military and surveillance use of tech by Israel

Time: 1 year total, with 1 month between Guardian revelations and Microsoft policy change, though the wider campaign had been going longer

Frequency: 2 protests reported

Location: US

Agent type: Employees, including tech workers and the campaign group No Azure For Apartheid

Type of activity: Disruption through sit-ins and disruptive protest/comments

Motivation for anti-AI: Employees did not want their work used for surveillance or military targeting of Palestinians

Outcome: Partial success. Microsoft announced it would disable some services.

• Microsoft storing Israeli intelligence trove used to attack Palestinians
• Update on ongoing Microsoft review
• ‘A million calls an hour’: Israel relying on Microsoft cloud for expansive surveillance of Palestinians
• Microsoft AI CEO’s remarks interrupted by pro-Palestinian protester
• No Azure for Apartheid

Sector: Pause AI campaign

Issue: Military use

Grievance: OpenAI retracted its ban on military work

Time: 2024

Frequency: 1 (Feb 2024)

Location: San Francisco

Agent type: Campaign group, anti-AGI activists with Pause AI

Type of activity: Demonstration

Motivation for anti-AI: Being against AGI/militarisation – no sign of an end to military use for OpenAI – in 2025 OpenAI wins a major military contract – and more here  on links to Pentagon 

Outcome: No visible impact

• Source 1
• Source 2
• Source 3
• Source 4

Sector: Tech

Issue: AI use in work (AI coding)

Grievance: Opposition to forced use of AI to code

Time: Around 1 month

Frequency: Ongoing for around one month

Location: US

Agent type: Employees (2 software engineers at Mixus)

Type of activity: Footdragging through refusal to use Cursor or other AI tools to code

Motivation for anti-AI: Lack of confidence that AI would improve work, and concern for deskilling

Outcome: No visible impact. The coders were fired, although this reflects a broader issue in the coding sector.

• Rise of the Cursor Resistance: Why Some Techies Want to Ignore AI Coding Tools
• AI Coding Resistance at Work: A Look at the Growing Disconnect and Ways to Ease the Stress

Sector: Screenwriting

Issue: AI use in work (for screenwriting)

Grievance: Potential job displacement and threats to creativity

Time: Around 5 months (source)

Frequency: One strike action lasting 146 days

Location: US

Agent type: Artists: screenwriters organised in the Writers Guild of America

Type of activity: Strike

Motivation for anti-AI: deskilling, job displacement, concerns over plagiarism/copyright – some picketline discussions reported here and a good summary here.

Outcome: Successful. the new contract included significant guardrails including ““cannot use AI to write scripts or to edit scripts that have already been written by a writer”, Conover says. The contract also prevents studios from treating AI-generated content as “source material”” source and WGA release on outcomes here.

• Source 1
• Source 2
• Source 3
• Source 4
• Source 5

Sector: Local government

Issue: AI use in work – government services

Grievance: Risks to job displacement and deskilling

Time: 2025 (3 months)

Frequency: No single event; regular negotiations

Location: US

Agent type: Employees (members of union SEIU local 668)

Type of activity: Union negotiations (workplace negotiations)

Motivation for anti-AI: Concerns over job displacement and deskilling

Outcome: Partial success. There was no agreement to avoid job losses, but a worker board was created to oversee the introduction of AI.

• Source 1
• Source 2

Sector: Healthcare

Issue: AI use in work – use of AI to provide healthcare by firms like Hippocratic AI and Xoltar

Grievance: Job displacement, deskilling, and safety risks

Time: 2025, 12 months

Frequency: 20 reported

Location: US

Agent type: Employees (nurses) organised by National Nurses United

Type of activity: Demonstrations

Motivation for anti-AI:  job displacement, deskilling, safety risks – concerns over inaccurate/unreliable health care provision from AI, the threat to human roles due to automating nursing provision, plus deprofessionalisation and disciplinary measures for employees who refuse to use AI

Outcome: Limited impact. There has been no obvious shift in approach by healthcare providers, but there has been some move toward regulation in California (AB1979).

• Source 1
• Source 2
• Source 3

Sector: Healthcare

Issue: AI use in work – use of AI to provide healthcare – seek guardrails as part of contract negotiation

Grievance: Job displacement, deskilling, and safety risks

Time: 2026, 2 months strike action

Frequency: One strike for around 2 months ongoing

Location: US (New York)

Agent type: Employees (nurses) organised by New York State Nurses Association (NYSNA)

Type of activity: Strike

Motivation for anti-AI: deskilling/safety – concerns over inaccurate/unreliable health care provision from AI, the threat to human roles due to automating nursing provision, plus deprofessionalisation and disciplinary measures for employees who refuse to use AI

Outcome:  successful – all strikes settled with agreements on safeguards (some details – commitment to keep real nurses at bedsides)

• Source 1
• Source 2
• Source 3
• Source 4
• Source 5
• Source 6

Sector: Healthcare

Issue:  AI use in work (AI software used for reporting)

Grievance:  job displacement – A pathway to job loss – latest update is here (Feb 2026) – union position is “A.I. used to assist therapists, not replace” which KP do not agree to.

Time:  initial complaint filed July 2025 – March 2026 (10 months)

Frequency: 1 strike day (march 2026)

Location: California

Agent type: Employees (nurses) organised by NUHW (IBHS branch)

Type of activity: strike – ongoing contract negotiations (for general nurses)

Motivation for anti-AI: job displacement

Outcome: incomplete – dispute ongoing (as of Feb 2026)

• Mental health clinicians hold major strike with support from Kaiser Permanente nurses, engineers

Sector: Film and TV industry

Issue: AI use in work (against AI writing script, and job losses)

Grievance: job displacement – IP, job loss

Time: July – Nov 2023 (5 months)

Frequency: 1 single strike action (for 5 months)

Location: California

Agent type: artists (Actors  – more details: 160,000 actors, musicians, dancers, announcers, voiceover artists, and stunt performers)

Type of activity: strike

Motivation for anti-AI:  Job loss

Outcome: Successful

• The actors’ strike is all but over. What do we know about the deal?
• Would You Do It Again? A Year After Strikes, Hollywood Reckons With the Aftermath
• Dis/engaging the ‘common sense’ of AI: Labor strategies from the 2023 SAG-AFTRA around data-driven technologies

Sector: Transport industry

Issue: AI use in work (job loss – getting power to veto AV deployment)

Grievance:  job displacement (job loss and road users safety)

Time: March 2024 (1 month)

Frequency:  n/a

Location: Ohio

Agent type: employees – Union (TWU)

Type of activity: union negotiations (contract negotiation)

Motivation for anti-AI:  job displacement and road safety

Outcome: Successful – the outcome of the union-management negotiations was for any replacement of bus drivers with autonomous vehicles requires union sign-off

• TWU Tech Newsletter: Winning Contracts That Fight Autonomous Technology
• On a drivers’ union victory against autonomous bus technology being deployed in Ohio, USA

Sector: Transport industry

Issue: AI use in work (job loss – getting power to veto AV deployment)

Grievance:  job displacement (job loss and road users safety)

Time: March 2024 (1 month)

Frequency:  n/a

Location: Ohio

Agent type: employees – Union (TWU)

Type of activity: union negotiations (contract negotiation)

Motivation for anti-AI:  job displacement and road safety

Outcome: Successful – the outcome of the union-management negotiations was for any replacement of bus drivers with autonomous vehicles requires union sign-off

• TWU Tech Newsletter: Winning Contracts That Fight Autonomous Technology
• On a drivers’ union victory against autonomous bus technology being deployed in Ohio, USA

Sector: Film and television

Issue: legal protection for the usage of AI

Grievance:  job displacement

Time: 2024 (1 month)

Frequency:  n/a

Location: Hollywood, US

Agent type: Employees – Union (IATSE)

Type of activity: union negotiations – contract negotiation (with threat of another Hollywod strike in the background, and with support from actors)

Motivation for anti-AI: job displacement

Outcome: successful – agreement provides protection for the employee if they use AI tech in their work, the liability of any event is on the company, as they have to make sure that the employee use it according to the law, and the employee cannot be held liable for the misuse. This industry uses different types of tech already for the production – also guarantees regarding job displacement

• Film and TV crew members reach deal with Hollywood studios to avert strike
• 2024 IATSE Basic Agreement MOA

Sector: Entertainment / Arts

Issue: AI use in work – Use of AI on actor recruitment

Grievance: job displacement – Actors typically have their appearance scanned when filming starts. Due to worries that this might feed generative AI in future, actors wish to refuse to be scanned.

Time: December 2025

Frequency: 1 time vote

Location: UK

Agent type: artists – actors, represented by union (Equity)

Type of activity: job displacement – prevention of scanning of appearance for AI use.

Motivation for anti-AI: job displacement

Outcome: incomplete – indicative vote passed with 99.6% of respondents voting yes to industrial action over AI and 71.5% of respondents voting that they would agree to refuse to be scanned on set – negotiations ongoing (as of Apr 2026)

• UK actors vote to refuse to be digitally scanned in pushback against AI
• Indicative ballot for AI protections

Sector: Data centres

Issue: Residents oppose new data centre (Southside Network Authority development)

Grievance: Local environment, especially noise pollution, plus water and energy use, construction, and traffic concerns

Time: 2025, c. 2 months between initial proposal becoming public knowledge and the vote by council to reject the plan

Frequency: Ongoing campaign and petition, emails, meeting attendance, and signs

Location: US

Agent type: Residents

Type of activity: Publicity campaign (petitions, emails, public hearings, and ‘no data centre’ signs)

Motivation for anti-AI: Noise from data centres

Outcome: Success. The council unanimously voted to block the proposal.

• Source 1
• Source 2
• Source 3

Sector: Data centres

Issue: Oppose new very large data centre in Bessemer, Alabama — Project Marvel (18 buildings, $14b) proposed by Logistics Land Investments LLC

Grievance: water supply, energy use, noise pollution, light pollution, deforestation (and loss of wildlife that lives there), pollution of local rivers/lakes, energy costs and secrecy (NDA conceals most details, including who the end-user is)

Time: 1 year from early 2025 to Feb 2026 (ongoing)

Frequency: Ongoing campaign

Location: US

Agent type: residents (supported by Alabama Rivers Alliance and the Southern Environmental Law Center and Center for Biological Diversity)

Type of activity: civic campaign including attend public hearings (wearing red t-shirts to indicate opposition), legal challenge on consultation process, petition for Birmingham darter, a fish newly discovered in one of the land’s creeks, in lake to be classified as endangered species

Motivation for anti-AI: Impact on the local area, pollution, and environmental damage

Outcome: Partial success. in stalling the plan so far – postponed initial decision due to judge ruling that the consultation was insufficient, then in Nov 2025 Council votes to approve, Jan 2026 the Council announced changes to the plan (to buy more land so to create a larger buffer between the data centre and local residents – and so as to avoid planned highway construction which had already been approved/planned in same place) – this means data centre is moving and so requires a second council decision on this new rezoning application.

• Source 1
• Source 2
• Source 3
• Source 4
• Source 5
• Source 6

Sector: Data centre

Issue: Oppose “hyperscale” data centre in Wilsonville, Alabama (14 buildings) proposed by NorthPoint Development

Grievance: Damage to the local environment

Time: August 2025 to Feb 2026 (ongoing)

Frequency: Intermittent reports of activity

Location: US

Agent type: Residents

Type of activity: Civic campaign — attendance at a town hall with signs saying ‘no data centre’

Motivation for anti-AI: Damage to the local environment

Outcome: Partial success. in stalling the decision so far and in Jan 2026 rumours the developer was thinking of relocating to a nearby area where rezoning had been approved.

• Source 1
• Source 2
• Source 3

Sector: Data centres

Issue: Data centres

Grievance: Local environment (impact of data centres on environment)

Time: 2025 (3 months to Feb 2026)

Frequency: N/A

Location: US

Agent type: NGOs

Type of activity: Online statement

Motivation for anti-AI: Environmental impact of data centres

Outcome: Partial success. The issue of moratoriums is on the agenda (although not obvious this is due to this call from environment groups)  – see South Dakota and New York –  by March 2026 11 states were considering legislation and Maine is on track to be first state to ban new data centres – New York State has adopted a moratorium (June 2026), as have now several other cities – only partial success as this is (as of Apr 2026) being done on city-by-city and state-by-state basis rather than US-wide.

• Source 1
• Source 2
• Source 3
• Source 4
• Source 5
• Source 6
• Source 7
• Source 8

Sector: Data centres

Issue: Data centres

Grievance: Utility bills, water

Time: Dec 2025 (3 months to Feb 2026)

Frequency: N/A

Location: US

Agent type: Residents (local residents, plus farmers, environmentalists)

Type of activity: Attend council meeting – civic action

Motivation for anti-AI: Damage to local environment, impact on bills

Outcome: Successful – rezoning application denied in Feb 2026.

• Source 1
• Source 2
• Source 3
• Source 4
• Source 5

Sector: Data centres

Issue: Local environment, utility costs, noise, light

Grievance: Local environment, utility costs, noise, light

Time: Roughly summer 2025 to Nov 2025

Frequency: N/A

Location: US

Agent type: Residents

Type of activity: civil campaign, including petition and general resident opposition, including attending council meeting.

Motivation for anti-AI: Local environment

Outcome:  Successful – planning committee agreed not to recommend rezoning; then the appeal hit legal problems in Nov 2025 eventually prompting developer to withdraw the proposal.

• Source 1
• Source 2
• Source 3
• Source 4
• Source 5
• Source 6

Sector: Data centres

Issue: Data centres (Project Blue, Amazon)

Grievance: Local environment, water use, distrust in Big Tech

Time: June – August 2025

Frequency: At least 3 public meetings

Location: US

Agent type:  Residents (supported by No Desert Data Center Coalition)

Type of activity: Attend public council meetings (one of up to 1000 members)

Motivation for anti-AI: Local environment, water, distrust of Big Tech

Outcome: Partial success – partial success – Council voted unanimously against in Aug 2025, and Amazon withdraws altogether in Dec 2025 – but project (relocated to Marana Town) gets approval in December 2025 (without an end user lined up) – Project Blue itself seems set to get final approval for two other sites anyway (Pina County and Marana Town) – also note the unions supported the proposal on basis of securing jobs.

• Source 1
• Source 2
• Source 3
• Source 4
• Source 5
• Source 6
• Source 7

Sector: Data centre

Issue: Data centre

Grievance: Local environment (risks to water, electricity supply)

Time: 3 months up to Sep 2025

Frequency: N/A

Location: US

Agent type: Residents (Protect Franklin Township)

Type of activity: demonstration – “around 60 protestors gathered outside a meeting on September 8” – civic campaign, including demonstration outside the county hall

Motivation for anti-AI: Local environment, water, electricity

Outcome: Successful – Google eventually withdrew the application

• Source 1
• Source 2

Issue: Data centres

Grievance: Local environment (opposition to the redevelopment plan and impact of the data centre on the local environment, plus gentrification)

Time: 2024 (30 months to Feb 2026)

Location: East London

Agent type: Residents

Type of activity: civic campaign: Letter writing, legal (taking a case to the Supreme Court) – although these are all related to the shopping centre plans – no details of campaigns specific to the data centre (since Apr 2026 we have seen several public demonstrations) – Petition  – Lobbying the GLA and central government’s inquiry, input into inquiry.

Motivation for anti-AI: Stopping replacing Truman Brewery as a landmark with data centres

Outcome: Incomplete – no decision from the government yet

• Source 1
• Source 2
• Source 3
• Source 4
• Source 5
• Source 6

Sector: Data centres

Issue: Data centres

Grievance: Environment and water usage, plus electricity

Time: Jan 2025 (13 months to Feb 2026)

Frequency: n/a

Location: Coweta, Oklahoma, US

Agent type: Residents

Type of activity: Publicity campaign, including a petition, publicity work, and a Facebook group

Motivation for anti-AI: Local environment, with concern over the impact of building data centres

Outcome: Partial success so far. The planning body already rejected recommending the proposal, but the council votes had not yet been cast as of Apr 2026.

• Source 1
• Source 2
• Source 3

Sector: Data centres

Issue: Data centre

Grievance: Environment

Time: 2025/2026 (4 months)

Frequency: n/a

Location: Oklahoma

Agent type: residents –  petition attend council meeting

Type of activity: Publicity campaign using signs, stickers, T-shirts, and printed flyers, plus a petition and attendance at council meetings

Motivation for anti-AI: Preserving community

Outcome: No visible impact – Feb 2026 the planning was approved (although legal challenge is ongoing)

• Source 1
• Source 2
• Source 3
• Source 4

Sector: Data centres

Issue: Data centre

Grievance: Local environment, including environmental fears, questions about conflicts of interest, and long-term impacts on children and property values

Time: 2025 (4 months)

Location: Archbald, Pennsylvania, US

Agent type: Residents

Type of activity: Civic campaign pressuring the council to reject the zoning proposal for data centres

Motivation for anti-AI: Local environment

Outcome: No visible impact. The council voted despite the public pressure.

• Source 1

Sector: Data centres

Issue: Data centres

Grievance: Local environment

Time: 2025 (4 months to Feb 2026)

Frequency: n/a

Location: US – Page City, Arizona

Agent type: Residents

Type of activity: Civic campaign: Campaign pressure for council to put this on public vote-  Petitioning, recall effort of the Mayor, seek referendum, protest/demonstration.

Motivation for anti-AI: Local environment

Outcome: Incomplete. No final decision had been made as of Apr 2026. A referendum petition was rejected on a spurious technicality, and the rezoning decision was still pending.

• Source 1
• Source 2
• Source 3
• Source 4

Sector: Data centres

Issue: Data centres

Grievance: Local environment

Time: 2025 (6 months to Feb 2026)

Frequency: n/a

Location: US – St. Joseph County, Indiana

Agent type: Residents

Type of activity: Civic campaign, including attending council meetings

Motivation for anti-AI: They already had big tech data centres in their community, and this campaign opposed new ones

Outcome: Successful. The proposal was rejected by council vote.

• Source 1
• Source 2

Sector: Data centres

Issue: Data centre (Landover)

Grievance: Local environment, including impacts on air, water, and woodland quality, and on quality of life in surrounding areas

Time: Autumn 2025 (6 months to Feb 2026)

Frequency: n/a

Location: Prince George’s County, Maryland, USA

Agent type: Residents, led by the Socialist Party

Type of activity: Publicity and civic campaign, including petitions with over 20,000 signatures, canvassing, lobbying, and campaigning

Motivation for anti-AI: Local environment

Outcome: Partially successful, after the pushback, there was a task force (and pause) to review and provide recommendations over the Landover data centre project – task force reported in Jan 2026 and looks set to extend the pause until April 2026 (update: appears to be stalled now until at least Sep 2026 and possibly for 2 more years – source).

• Source 1
• Source 2
• Source 3
• Source 4

Sector: Data centres

Issue: Data centre

Grievance: Local environment

Time: Oct 2024 to Apr 2025 (7 months)

Frequency: n/a

Location: US

Agent type: Residents, supported by Southern Environmental Law Center

Type of activity: Civic campaign, including attending council meetings, holding meetings, and lobbying local government

Motivation for anti-AI: Local environment

Outcome: Successful. Pittsylvania’s local government rejected Balico’s request to rezone the land for data centers in Apr 2025.

• Source 1
• Source 2

Sector: Data centres

Issue: Data centre

Grievance: Local environment

Time: Feb 2025 to Nov 2025 (7 months)

Frequency: n/a

Location: US

Agent type: Residents

Type of activity: Publicity campaign and demonstrations

Motivation for anti-AI: Local environment

Outcome: No visible impact. The council was unable to act as of Apr 2026.

• Source 1
• Source 2
• Source 3
• Source 4

Sector: Data centres

Issue: Data centre

Grievance: Local environment

Time: 2024 (18 months to Feb 2026)

Frequency: n/a

Location: Chile

Agent type: Residents

Type of activity: publicity campaign –  demonstrations, legal actions,Political Lobbying and Referendums, Dialogue with Corporations, and data gathering (research on the impact in the community), local campaign

Motivation for anti-AI: Local environment

Outcome: Successful (Google data centre permit is pulled, 2024)

• Google takes Chile data center plans back to square one on environmental concerns
• Chile partially pulls Google data center permit, seeks tougher environmental checks
• Chile: Tech giants build dozens of data centers while activists raise environmental concerns

Sector: Data centres

Issue: Data centre

Grievance: Local environment (water)

Time: 2025 (9 months to Feb 2026)

Frequency: n/a

Location: Chile

Agent type: Residents

Type of activity: publicity campaign –  demonstrations, legal actions,Political Lobbying and Referendums, Dialogue with Corporations, and data gathering (research on the impact in the community), local campaign

Motivation for anti-AI: Local environment

Outcome: no visible impact (Amazon 2025, given the environmental clearance to go ahead)

• U.S tech giants are building dozens of data centers in Chile. Locals are fighting back

Sector: Data centres

Issue: Data centre

Grievance: Local environment (water)

Time: 2023 (30 months to Feb 2026)

Frequency: n/a

Location: Chile

Agent type: Residents

Type of activity: publicity campaign –  demonstrations, legal actions,Political Lobbying and Referendums, Dialogue with Corporations, and data gathering (research on the impact in the community), local campaign

Motivation for anti-AI: Local environment

Outcome: no visible impact (Amazon 2025, given the environmental clearance to go ahead)

• Chile: Tech giants build dozens of data centers while activists raise environmental concerns

Sector: Data centres

Issue: Data centre

Grievance: local environment – Electricity usage, Water consumption, flawed consultation

Time: Sep2025 (6 months to Feb 2026)

Frequency: n/a

Location: Brazilian state of Ceará

Agent type: Indigenous community (Indigenous group in the Brazilian state of Ceará, in the country’s northeast (Anacé Indigenous community)

Type of activity: lawsuit (Filing complaint and legal action in court)

Motivation for anti-AI: Local environment

Outcome: Incomplete (as of Feb 2026)

• Indigenous group in Brazil takes TikTok to court over planned data center
  

Sector: Data centres

Issue: Data centre

Grievance: local environment – Energy consumption, air quality, noise, lack of transparency in decision-making (approving a contract for data centre deployment without council approval by former Mayor)

Time: Jan 2026

Frequency: n/a

Location: Colombiana Alabama

Agent type: residents

Type of activity: civic campaign – campaigning through council

Motivation for anti-AI: Local environment

Outcome: limited impact – data centre went ahead however, they managed to push the company to amend some of the details of the agreement, but overall they did not manage to prevent it as the contract was secretly secured by the former mayor – limited concessions (“DigiPowerX agreed to switch to a glycol-cooled, closed-loop system after the new administration made clear that the town of around 4,500 did not have the capacity to supply the company with industrial amounts of water.”)

• An Alabama Mayor Signed an NDA With a Data Center Developer

Sector: Advertising

Issue: AI-generated advertisement

Grievance: AI output, unsettling AI art

Time: December 2025, 1 month

Frequency: Consistent online criticism through December 2025

Location: Netherlands

Agent type: Online public

Type of activity: Online criticism (Online lambasting)

Motivation for anti-AI: Dislike of AI-generated content

Outcome: Success. The advert was pulled.

• Source 1
• Source 2

Sector: Porn

Issue: Deepfake images

Grievance: Sexual harassment and abuse through non-consensual deepfake porn

Time: 2024 to 2025 (around 12 months)

Frequency: Consistent lobbying and activism

Location: UK

Agent type: Campaign group, End Violence Against Women

Type of activity: Civic campaign, including direct lobbying of policymakers

Motivation for anti-AI: Regulation to ban deepfake porn

Outcome: Successful. The Data (Use and Access) Act was adopted in 2025.

• Source 1
• Source 2

Sector: Social media

Issue: deep fake images (sexual harassment/abuse/images – Inappropriate outputs from generative AI)

Grievance: Chatbot producing unethical and illegal content

Time: End of 2025 – beginning of 2026

Frequency: Continued online commentary and official investigations

Location: global – US company, global condemnation

Agent type: Online public – individuals, regulatory bodies

Type of activity: online criticism – Negative online commentary, official investigations

Motivation for anti-AI: AI as source of sexual harassment/abuse/images – aim to prevent chatbots from producing unethical content

Outcome: Successful. Widespread condemnation and introduction of safeguard and prohibitive measures on a global scale (including EU, India, the UK, Malaysia, Indonesia, and Brazil, amongst others.

• Grok produces sexualized photos of women and minors for users on X – a legal scholar explains why it’s happening and what can be done
• AI Act: deal on simplification measures, ban on “nudifier” apps
• India gives Musk 72 hours as X floods with obscene images generated using Grok
• What to know about UK legal changes aiming to regulate AI-generated nude images
• Malaysia and Indonesia Block Access to Grok Because of Sexually Explicit Content
• Brazil Tells xAI To Remove Sexualised Grok Content

Sector: Employment

Issue: Discrimination

Grievance: AI discrimination in the employment process

Time: filed 2023 – ongoing (36 months to Feb 2026)

Frequency: Not applicable

Location: USA, California

Agent type: Employees (job applicant)

Type of activity: Lawsuit

Motivation for anti-AI: Discrimination by age, with the claim that the algorithm never let the applicant get through the first round of applications

Outcome: Incomplete. No ruling yet as of Apr 2026, but the case was allowed to proceed as a collective action in May 2025, and in March 2026 the court rejected Workday’s argument that hiring was not covered by discrimination legislation.

• SocialTalent summary of the collective action ruling
• Columbia Black Pre-Law analysis
• Washington Post coverage
• Forbes update on the opt-in process

Sector: Business / tech

Issue: Agentic AI – unauthorised use of AI on platform

Grievance: Covert use of agentic AI

Time: November 2025

Frequency: Not applicable

Location: US

Agent type: Corporation (Amazon)

Type of activity: Lawsuit

Motivation for anti-AI: Non-disclosure of agentic AI status, with Amazon arguing that Perplexity should declare when its agent is shopping on its website

Outcome: incomplete (as of Apr 2026) – still to be decided – In March 2026, Amazon managed to get a temporary block on perplexity whilst it was waiting for the judgement, but this was in the same month overturned on appeal – details here.

• TechStartups report
• Archived coverage
• JD Supra legal summary
• Reuters update on the temporary block and appeal

Sector: Software

Issue: Poor quality

Grievance: Poor quality

Time: 2025 (December 2025 to Feb 2026, 3 months)

Frequency: Ongoing online criticism and mocking

Location: Global (online)

Agent type: Customers/users

Type of activity: Online criticism and mocking

Motivation for anti-AI: Poor quality

Outcome: limited impact –  seems unpopular and low take-up rate, but doesn’t seem to be due to the online criticism in any way (rather the other way around)

Low use rate – although that seems to be more to do with the quality of the model, rather than the online mocking – “active Copilot users only account for a measly 2.5% of all Windows 10 and 11 user”

Microsoft AI CEO, Mustafa Suleyman, responds defensively to the criticism 

• Substack commentary
• Reddit discussion
• Quartz coverage
• Windows Central report
• BBC coverage
• Perspectives analysis

Sector: Software

Issue: Poor quality

Grievance: Community safety and public transport

Time: since 2023

Frequency: unknown

Location: San Francisco

Agent type: road activists

Type of activity: direct action, confusing autonomous car algorithms with the use of cones

Motivation for anti-AI:  less car-dominant San Francisco, promoting public transportation and advocating for pedestrian safety.

Outcome: Limited impact. limited impact – initially successful sabotage, caused disruption and gained media attention – overall though autonomous cars became commonplace in San Francisco.

• TheGuardian

Sector: Politics

Issue: far right politics

Grievance: ChatGPT is Trump’s biggest donor and is used by ICE

Time: January 2026

Frequency: Ongoing boycott, regular online and media presence encouraging boycott.

Location: USA (but global boycott encouraged)

Agent type: customers/users – Grassroots organisers

Type of activity: Boycott

Motivation for anti-AI: Encouraging people to boycott ChatGPT to damage their business, encourage AI companies to prevent fascists from using their tech.

Outcome: partial success – getting increasing attention.

• https://quitgpt.org
• A QuitGPT campaign is urging people to cancel chatgpt subscriptions

Sector: Politics

Issue: Semiconductor industry

Grievance: local environment – Following pressure from clients such as Apple and AMD to decarbonise, Taiwan is aggressively expanding renewable energy. Offshore windfarms are having a negative on ecosystem, agriculture, and fishing.

Time: Since 2023 (30 months)

Frequency:

Location: Taiwan

Agent type: local fishers – Local residents, farmers, and fishers.

Type of activity: Demonstrations

Motivation for anti-AI: Preventing more wind farms from being built.

Outcome: no visible impact – Expansion of wind farms still planned – seems to be going ahead as last commissioned plant installed Jan 2026. Note that this report (March 2025) mentions lengthy delays due in part to local opposition – but this seems to be a different installation (Yunlin, not Fangyuan, Changhua county)

• AI’s green-energy goal is devastating Taiwan’s coastal villages
• Central Taiwan Residents Want Wind Farm Project Out of Their Backyard｜TaiwanPlus News
• All wind turbines installed at Ørsted’s 920 MW Greater Changhua 2b and 4 offshore wind farms in Taiwan

Sector: Museums / curation

Issue: AI output (AI generated content)

Grievance: mis/dis-information – Use of AI not transparent. Worries about blurring the line between historical evidence and digital fabrication.

Time: January 2026

Frequency: One-off bout of online backlash

Location: UK

Agent type: online experts – Archaeologists, people online

Type of activity: Online criticism

Motivation for anti-AI: mis/disinformation

Outcome: Successful. Posts removed.

• British Museum Faces Backlash Over AI-Generated Posts