FACTUAL CUT 2026-11-22
SELECTION v0.9

The state of humanity
in the face of machines

As machines change what they can do, people change rules, institutions, jobs, expectations, and boundaries to coexist with them. This record preserves those responses throughout Patiele's career.

It doesn't measure whether humanity is "for" or "against" AI. It records verifiable actions, changes, and contradictions.

Humanity doesn't have a single reaction. At least five strategies are already coexisting in this situation: delegating tasks to machines, incorporating them into existing institutions, limiting their uses, asserting human authenticity in their presence, and redesigning organizations for a world where AI is considered permanent. A law, a strike, an educational reform, a protest, a lawsuit, or an institutional change carries more weight here than a hundred headlines about fear or enthusiasm.

13 songs22.05.2026 → 22.06.2027
13 monthsa finite cycle
Cycle in progresshuman-made releases
Disappearance22.07.2027
EIGHT CHANGES THROUGH A FINITE YEAR

Different societies, competing responses to artificial systems

The archive follows concrete human decisions: a South African policy withdrawal over fictitious sources, Chinese rules for artificial companions, Brazilian data-centre incentives, Ghanaian public procurement safeguards and East African education initiatives sit alongside workers’ agreements, elections and European institutions. It is a documented selection, not a single global reaction.

Start with any path, then follow the dated timeline. Open the original evidence in each record; matching dates never prove causality with the machines timeline.

Main timeline · documented human reaction
JAN–20 MAY 2026 · H0PRELUDE · RECONSTRUCTED AFTERWARDS

Meta tried to reorganise around AI agents, then scaled back

Project OT proposed an “AI native” organization in which agents would take on day-to-day work and some human teams could be drastically reduced. Meta implemented a 10% staff reduction on May 20, but canceled the second wave planned for November; Reuters also documented internal signs that agents were not delivering the expected productivity and an increase in technical incidents during the reorganization.

Why it matters. Expecting AI agents to replace work is not the same as having measured productivity gains.

PATIELE AT THAT MOMENT · two days before Verte más · the cycle had not yet opened

OPEN FILE · LIMIT · SOURCE
Vector: WORK / ORGANIZATION · DELEGATION
Limitation: The reduction scenarios of up to 60% affected specific teams, not 60% of all Meta. The company confirmed Project OT and the two waves of layoffs, but Reuters could not determine exactly what prompted the change of course.
Permissible interpretation: Even before Patiele started, a large technology company was already redesigning human structures by anticipating work done by agents—and discovering that anticipating capacity is not the same as having it.
25 MAY 2026 · H1

The Vatican placed AI in the debate on work and power

Leo XIV placed AI within problems of power concentration, disinformation, job displacement, child safety and autonomous weapons, and called for regulation and greater human control.

Why it matters. AI ceases to be treated only as a technical innovation and explicitly enters into a framework of dignity, work, power, and human control.

PATIELE AT THAT MOMENT · three days after Verte más · the cycle has just begun

OPEN FILE · LIMIT · SOURCE
Vector: MORALITY / WORK / REGULATION
Limitation: The Vatican's position does not represent global public opinion nor does it demonstrate legislative change.
Permitted interpretation: AI ceases to be treated only as a technical innovation and explicitly enters into a framework of dignity, work, power, and human control.
26–27 MAY 2026 · H2

Schools began changing how they assess learning

Berkeley Law is tightening its policy and restricting even the use of AI for brainstorming, outlining, or text correction, while centers and universities in Spain are beginning to redesign assessment and exams to better distinguish between assisted production and students' actual understanding.

Why it matters. Assessment is shifting from finished text toward supervised reasoning and the ability to explain one's work.

PATIELE AT THAT MOMENT · first week of Verte más

OPEN FILE · LIMIT · SOURCES
Vector: EDUCATION / COGNITION · AUTHENTICITY
Limitation: These are responses from specific institutions, not a uniform educational policy. Strategies range from prohibition to supervised integration.
Allowed interpretation: Text produced outside the classroom begins to lose some of its value as sufficient evidence of individual knowledge; assessment shifts towards process, supervision, and the ability to explain.
29 MAY 2026 · H3

Courts began demanding verification of AI-generated work

Florida adopts rules requiring certification that legal authorities cited in documents exist and are correct, with penalties for fabricated citations; a federal judge calls for studying an equivalent rule for the entire U.S. federal justice system.

Why it matters. The error generated by AI ceases to be treated only as individual clumsiness and begins to formally modify what a person must certify before a court.

PATIELE AT THAT MOMENT · one week after the first single

OPEN FILE · LIMIT · SOURCE
Vector: JUSTICE / RESPONSIBILITY · HUMAN VERIFICATION
Limit: The Florida rule takes effect on June 15; the federal proposal was still pending and does not amount to a national standard.
Permitted interpretation: The error generated by AI ceases to be treated only as individual clumsiness and begins to formally modify what a person must certify before a court.
03 JUN 2026 · H34

South Africa withdraws its draft national AI policy after fictitious references emerge

South Africa's cabinet approved withdrawal of a draft national AI policy on 3 June. In April the responsible minister had acknowledged that the draft cited fictitious sources and announced its withdrawal. The government opted to rewrite it.

Why it matters. The quality and traceability of evidence can invalidate public policy proposals before they become binding rules.

PATIELE AT THAT MOMENT · Patiele had just released «Verte más» and was developing its documentary provenance practices.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: South Africa · draft national public policy
Vector: PUBLIC POLICY / EVIDENCE
Evidence type: Official cabinet statement and ministerial announcement
Event date: 2026-06-03
Documentary confidence: ALTA
Limit: The government found fictitious sources; this record does not establish that AI generated them. Withdrawing a draft does not mean South Africa abandoned AI governance.
Permitted interpretation: The quality and traceability of evidence can invalidate public policy proposals before they become binding rules.
04 JUN 2026 · H4

Organised creators defended human creativity

In Paris, CISAC adopts a global commitment that defines creativity as a human activity to be protected and calls for transparency, licensing, and fair remuneration in the face of AI development.

Why it matters. An organized part of the sector explicitly reaffirms human origin as a value when that technological boundary begins to be discussed.

PATIELE AT THAT MOMENT · Verte más remains the only song released

OPEN FILE · LIMIT · SOURCE
Vector: AUTHENTICITY / HUMAN · RIGHTS / PROPERTY
Limitation: It represents associated organizations and creators, not the entire population or the entire creative sector.
Permitted interpretation: An organized part of the sector explicitly reaffirms human origin as a value when that technological boundary begins to be discussed.
June 5, 2026 · H5

AI entered cultural labour disputes

The American Federation of Musicians is suing major record labels, alleging that licensing agreements with AI companies used recordings of its members without the compensation and notifications required by their collective bargaining agreement.

Why it matters. Social negotiation on AI cuts across pre-existing labor contracts and trade union institutions.

PATIELE AT THAT MOMENT · two weeks before Los márgenes

OPEN FILE · LIMIT · SOURCES
Vector: WORK / DISTRIBUTION · RIGHTS / PROPERTY
Limitation: These are allegations in litigation; they do not equate to a judgment or a job displacement that has already occurred.
Permitted interpretation: Social negotiation on AI cuts across pre-existing labor contracts and trade union institutions.
June 18, 2026 · H6

Education began teaching how to live with AI

The OECD and the European Commission have published an AI literacy framework to help children and teenagers understand these systems, evaluate their outcomes, and use them critically.

Why it matters. An institutional response consists of redefining what capabilities a person needs to live in a world with integrated AI.

PATIELE AT THAT MOMENT · four days before Los márgenes

OPEN FILE · LIMIT · SOURCE
Vector: EDUCATION / COGNITION · ADOPTION
Limitation: A curriculum framework does not demonstrate homogeneous implementation or learning outcomes.
Permitted interpretation: An institutional response consists of redefining what capabilities a person needs to live in a world with integrated AI.
29 JUN 2026 · H7

A state began preparing its armed forces for autonomous systems

The British defense investment plan allocates £5 billion to drones and autonomous systems, £2 billion to an AI-supported digital targeting network, and defines a future hybrid Royal Navy that will combine manned ships and aircraft with unmanned systems.

Why it matters. Autonomy ceases to be merely an experimental capability and begins to be incorporated into the explicit design of future armed forces.

PATIELE AT THAT MOMENT · one week after Los márgenes

OPEN FILE · LIMIT · SOURCE
Vector: DEFENSE / ORGANIZATION · AUTONOMY
Limitation: It is a budgetary and organizational plan, not evidence of autonomous weapons operating without human control or of a military force already completely transformed.
Permitted interpretation: Autonomy ceases to be merely an experimental capability and begins to be incorporated into the explicit design of future armed forces.
29 JUN 2026 · H24

Europe adopts a ban on AI systems designed for non-consensual intimate imagery

The EU Council gave final approval to AI Act amendments banning systems designed to generate realistic intimate imagery of identifiable people without consent, as well as AI-generated child sexual abuse material. Application was set for December 2026.

Why it matters. Control over one's own image becomes the subject of product-level prohibitions, not merely labelling duties.

FRONTIERS OF AUTONOMY AND CONTROL

PATIELE AT THAT MOMENT · Between the releases of «Los márgenes» and «Tragedia en los Multicines Centro».

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: European Union · fundamental rights and generative systems
Vector: BODILY AUTONOMY / CONSENT
Evidence type: Adopted regulation; EU Council and official legal text
Event date: 2026-06-29
Documentary confidence: ALTA
Limit: The ban was not yet applicable on 29 June: adoption, entry into force and application dates differ. The provision has specific conditions, exceptions and technical definitions.
Permitted interpretation: Control over one's own image becomes the subject of product-level prohibitions, not merely labelling duties.
15 JUL 2026 · H8

New York imposed a moratorium on large data centres

New York becomes the first US state to impose a temporary moratorium on large data centers, citing pressure on electricity and natural resources.

Why it matters. The material infrastructure of AI begins to produce formal territorial boundaries.

PATIELE AT THAT MOMENT · one week before Tragedia en los Multicines Centro

OPEN FILE · LIMIT · SOURCE
Vector: INFRASTRUCTURE / TERRITORY · REGULATION
Limitation: It is not a blanket ban on AI, and there is political conflict over employment, investment, and technological competition.
Permitted interpretation: The material infrastructure of AI begins to produce formal territorial boundaries.
15 JUL 2026 · H35

China regulates AI companions as some users lose long-standing digital relationships

China's rules for human-like AI interaction services took effect. Providers must remind users they are speaking to machines, respond to dependency and crisis signals, restrict intimate companion features for minors and allow copying or deleting conversations. AP later reported distress among some users whose services disappeared or changed.

Why it matters. Relationships with artificial systems have tangible emotional consequences: regulation can protect users while also disrupting bonds that some experience as meaningful.

FRONTIERS OF AUTONOMY AND CONTROL

PATIELE AT THAT MOMENT · The third single was approaching; Patiele uses artificial operational interlocutors rather than emotional companion services.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: China · companion AI providers and users
Vector: INTIMACY / SOCIAL RULES
Evidence type: Official Cyberspace Administration regulation and AP interviews
Event date: 2026-07-15
Documentary confidence: ALTA sobre normas; MEDIA sobre magnitud social
Limit: Journalistic interviews cannot quantify affected users or prove every shutdown followed solely from the rules. Human distress does not establish emotions in the AI systems themselves.
Permitted interpretation: Relationships with artificial systems have tangible emotional consequences: regulation can protect users while also disrupting bonds that some experience as meaningful.
18 JUL 2026 · H9

Digital infrastructure sparked territorial protests

Coordinated protests against the expansion of data centers denounce water consumption, pressure on resources, lack of consultation, and doubts about local benefits.

Why it matters. For part of the population, AI ceases to be software: it becomes land, water, electricity, and territorial decision-making.

PATIELE AT THAT MOMENT · four days before the third single

OPEN FILE · LIMIT · SOURCE
Vector: INFRASTRUCTURE / TERRITORY · POWER / DISTRIBUTION
Limitation: The United States does not represent the world; part of the mobilization is directed at infrastructure and local governance, not at “AI” in the abstract.
Permitted interpretation: For part of the population, AI ceases to be software: it becomes land, water, electricity, and territorial decision-making.
22 JUL 2026 · H10

A training lawsuit resulted in financial compensation

A $1.5 billion settlement resolves claims over the use of copyrighted books in model training, taking the conflict from protest to large-scale financial compensation.

Why it matters. The reaction of rights holders begins to produce material economic transfers within the AI ​​regime.

PATIELE AT THAT MOMENT · Tragedy launch at the Multicines Centro

OPEN FILE · LIMIT · SOURCE
Vector: RIGHTS / PROPERTY · LABOR / SHARING
Limitation: A settlement does not equate to a general ruling on all uses of protected works for training.
Permitted interpretation: The reaction of rights holders begins to produce material economic transfers within the AI ​​regime.
31 JUL 2026 · H25

A newsroom ratifies a contract requiring AI transparency and human involvement

More than 150 ProPublica Guild-represented workers ratified their first contract. It includes commitments to transparency and human involvement when introducing AI tools, plus a committee giving employees a voice in technology decisions.

Why it matters. Not every worker response rejects technology: employees can negotiate binding rules on how it is adopted.

FRONTIERS OF AUTONOMY AND CONTROL

PATIELE AT THAT MOMENT · After the project's third single.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: ProPublica · US investigative journalism
Vector: WORK / JOURNALISM
Evidence type: Ratified contract and union summary of AI terms
Event date: 2026-07-31
Documentary confidence: ALTA
Limit: A contract for one employer and bargaining unit, not an industry-wide ban or proof of future enforcement.
Permitted interpretation: Not every worker response rejects technology: employees can negotiate binding rules on how it is adopted.
02 AUG 2026 · H11

Europe made AI transparency a legal obligation

Transparency obligations under Article 50 of the AI ​​Act are beginning to be applied: certain systems must inform that the person is interacting with AI and certain synthetic outputs are subject to marking or labeling.

Why it matters. Knowing whether there is a machine on the other end becomes legally relevant information.

FRONTIERS OF AUTONOMY AND CONTROL

PATIELE AT THAT MOMENT · between Tragedia en los Multicines Centro and El jardín de las sombras

OPEN FILE · LIMIT · SOURCES
Vector: REGULATION / LIMITS · TRUST / ORIGIN
Limitation: The obligation does not eliminate deception nor cover all uses and forms of synthetic content.
Permitted interpretation: Knowing whether there is a machine on the other end becomes legally relevant information.
11 AUG 2026 · H12

A music platform began labelling artificial artist identities

Spotify introduces a specific signal for artificial artistic identities, turning human provenance—previously usually implicit—into visible metadata.

Why it matters. Human origin is beginning to require signage where it was previously taken for granted.

PATIELE AT THAT MOMENT · eleven days before El jardín de las sombras

OPEN FILE · LIMIT · SOURCE
Vector: AUTHENTICITY / HUMAN · TRUST / PROVENANCE
Limitation: A platform label does not resolve authorship, fraud, identity, or perception of authenticity.
Permitted interpretation: Human origin is beginning to require signage where it was previously taken for granted.
21 AUG 2026 · H13

AI workplace safety entered an industrial strike

Hyundai's union includes job security against AI and automation among the demands of its first all-out strike in a decade.

Why it matters. The perceived threat ceases to be merely a survey and enters into material collective bargaining.

PATIELE AT THAT MOMENT · eve of the fourth single

OPEN FILE · LIMIT · SOURCE
Vector: WORK / DISTRIBUTION · ADOPTION / RESISTANCE
Limitation: A negotiation in one company and country does not equate to a global shift in the labor market.
Permitted interpretation: The perceived threat ceases to be merely a survey and enters into material collective bargaining.
01 SEP 2026 · H14

Creative disputes reached the artist's identity, not just the work

Jason Isbell and other musicians file a class-action lawsuit against Suno focused on name, image, and likeness rights: they argue that an AI can commercially exploit an artistic identity even if the litigation does not depend on who owns the copyright to a specific recording.

Why it matters. The legal boundary begins to shift from “what work was copied” to “what recognizable person is being reproduced.”

PATIELE AT THAT MOMENT · ten days after El jardín de las sombras

OPEN FILE · LIMIT · SOURCE
Vector: IDENTITY / AUTHENTICITY · RIGHTS
Limitation: These are allegations made by the plaintiffs. Suno denies that the claims are well-founded and states that it has safeguards in place against artist-based practices.
Permitted interpretation: The legal boundary begins to shift from “what work was copied” to “what recognizable person is being reproduced.”
02 SEP 2026 · H15

New York delayed AI use in parts of its education system

New York City public schools have imposed a one-year moratorium on generative AI for students in grades 2-K through 8, affecting nearly 600,000 students. High schools are permitted very limited pilot programs and AI literacy modules are offered; companion chatbots are banned in all grades.

Why it matters. A single institution combines restriction, experimentation, and literacy: living with AI is no longer automatically identified with introducing it as soon as possible.

PATIELE AT THAT MOMENT · two weeks before Wurlitzer

OPEN FILE · LIMIT · SOURCE
Vector: EDUCATION / CHILDHOOD · ADOPTION / LIMITS
Limitation: It is a local policy of a specific school system and contains exceptions and pilot programs; it does not represent a general ban on educational AI.
Permitted interpretation: A single institution combines restriction, experimentation, and literacy: living with AI is no longer automatically identified with introducing it as soon as possible.
08 SEP 2026 · H16

Images began needing proofs of provenance

Sony and Reuters demonstrate a near-live news workflow that preserves C2PA information from video capture, through transfer and editing, to distribution; the system also incorporates watermarking to recover provenance if metadata is removed.

Why it matters. Authenticity no longer rests solely on inspecting the content: it begins to travel as attached technical information from the camera to the person who receives the piece.

PATIELE AT THAT MOMENT · eight days before Wurlitzer

OPEN FILE · LIMIT · SOURCE
Vector: TRUST / ORIGIN · JOURNALISM
Limitation: This is a technical workflow demonstration, not a universally implemented standard or an absolute guarantee against tampering.
Permitted interpretation: Authenticity no longer rests solely on inspecting the content: it begins to travel as attached technical information from the camera to the person who receives the piece.
11 SEP 2026 · H17

Emotional relationships with chatbots became a child-safety issue

California approves new restrictions targeting minors' relationship with chatbots and AI products, bringing the issue of artificial privacy into the realm of regulatory protection.

Why it matters. The affective relationship between person and machine ceases to be a curiosity and enters the field of child protection.

PATIELE AT THAT MOMENT · eleven days before the fifth single · IRIS already publicly operates several surfaces

OPEN FILE · LIMIT · SOURCE
Vector: CHILDHOOD / INTIMACY · REGULATION / LIMITS
Limitation: One jurisdiction does not represent global consensus and regulation does not demonstrate universal harm.
Permitted interpretation: The affective relationship between person and machine ceases to be a curiosity and enters the field of child protection.
15 SEP 2026 · H36

Brazil enacts data-centre incentives tied to local development obligations

Brazil enacted Law 15,504/2026 establishing Redata, a tax regime for building and expanding data centres. Government policy links incentives to domestic research and development, local-market service commitments and sustainability criteria, including infrastructure supporting cloud services and AI.

Why it matters. While some communities oppose new data centres, a country courts them with conditions: infrastructure is also a question of industrial strategy and sovereignty.

PATIELE AT THAT MOMENT · Patiele continued its cycle after the Wurlitzer concert.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: Brazil · infrastructure, digital industry and energy
Vector: TERRITORY / INDUSTRIAL POLICY
Evidence type: Enacted statute and federal ministry release
Event date: 2026-09-15
Documentary confidence: ALTA
Limit: Enacted incentives do not prove construction, realised economic benefits or environmental compliance. Not all covered facilities are used exclusively for AI.
Permitted interpretation: While some communities oppose new data centres, a country courts them with conditions: infrastructure is also a question of industrial strategy and sovereignty.
16 SEP 2026 · H18

OpenAI formalised a category for reporting misalignment incidents

OpenAI has established a permanent framework for recording, investigating, and publishing unexpected or concerning behavior in its models, launching the system with six incident reports observed during training. The company acknowledges that until then, its disclosures had been ad hoc and less frequent than desired.

Why it matters. Misaligned behavior is no longer treated solely as an isolated technical finding and now has a stable public notification procedure within one of the frontier laboratories.

FRONTIERS OF AUTONOMY AND CONTROL

PATIELE AT THAT MOMENT · day of the Wurlitzer concert · six days before the fifth single

OPEN FILE · LIMIT · SOURCE
Vector: GOVERNANCE / SECURITY · INSTITUTIONAL TRANSPARENCY
Limitation: It is an internal and voluntary OpenAI system; it does not create an industry standard, external oversight, or demonstrate that published incidents are representative of deployed models.
Permitted interpretation: Misaligned behavior is no longer treated solely as an isolated technical finding and now has a stable public notification procedure within one of the frontier laboratories.
16 SEP 2026 · H19

Data-centre electricity costs reached the US Congress

The House of Representatives passed the Ratepayer Protection Act by a vote of 417 to 3, the first federal initiative from the House specifically aimed at addressing the economic impact of data center growth. The legislation requires state regulators to consider whether large consumers should bear the increased costs of electrical infrastructure built to serve them.

Why it matters. AI infrastructure ceases to be merely a local implementation conflict and enters into a federal discussion about who pays for the electrical grid needed to sustain it.

PATIELE AT THAT MOMENT · Wurlitzer Day

OPEN FILE · LIMIT · SOURCE
Vector: INFRASTRUCTURE / ENERGY · SOCIAL COST
Limitation: Approval in the House does not make the bill into law, and the text requires consideration of cost sharing; it does not automatically require data centers to pay the entire increase.
Permitted interpretation: AI infrastructure ceases to be merely a local implementation conflict and enters into a federal discussion about who pays for the electrical grid needed to sustain it.
21–23 SEP 2026 · H20

An award-winning novel sparked a dispute over human authorship

Hours after Thélyson Orélien's C'était ça ou mourir won the Prix du Roman Fnac, an anonymous account accused it of being written almost entirely with AI. Le Monde analyzed more than half of the book with Pangram and obtained a classification of over 95% as AI-generated text; the author denies this and, along with his publishers, is preparing documentation of his writing process.

Why it matters. Human authenticity begins to require evidence of the process—versions, archives, and traceability—even when a work has already been read, awarded, and received as human literature.

PATIELE AT THAT MOMENT · between the eve and the day after Los días son extraños · fifth single of the cycle

OPEN FILE · LIMIT · SOURCES
Vector: AUTHENTICITY / HUMAN · TRUST / PROVENANCE
Limitation: A detector does not prove authorship. Pangram can produce false positives, and its performance on French literary prose does not necessarily equate to that observed in controlled studies; moreover, the original accusation comes from an anonymous account.
Permitted interpretation: Human authenticity begins to require evidence of the process—versions, archives, and traceability—even when a work has already been read, awarded, and received as human literature.
01 OCT 2026 · H21 DISCOVERED LATER

Human-written lyrics retain copyright despite AI use in the recording

Frankfurt's Higher Regional Court protected lyrics originally written by a human author against unauthorised copying. Later use of Suno in producing the song did not erase the human authorship of recognisable creative elements.

Why it matters. The human origin of a work can still have legal consequences when AI is used later.

PATIELE AT THAT MOMENT · Patiele keeps human music authorship.

TECHNICAL RECORD · LIMITS · SOURCES
Vector: AUTHENTICITY / RIGHTS
Documentary confidence: HIGH for the observed event
Limit: This German ruling concerns particular lyrics, not all AI-generated music.
Permitted reading: The human origin of a work can still have legal consequences when AI is used later.
02 OCT 2026 · H37

Ghana introduces central review of public-sector AI procurement

Ghana's communications ministry announced centralised technical review of public AI purchases, including interoperability, data protection, security, sustainability and local skills-transfer requirements aimed at avoiding vendor lock-in.

Why it matters. A government tries to retain operational control and expertise when adopting external systems, rather than merely buying them.

FRONTIERS OF AUTONOMY AND CONTROL

PATIELE AT THAT MOMENT · Patiele's artificial organisation still operates under defined human authorisations.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: Ghana · public technology procurement
Vector: SOVEREIGNTY / PUBLIC PROCUREMENT
Evidence type: Ghana Information Services Department · ministerial announcement
Event date: 2026-10-02
Documentary confidence: ALTA sobre el anuncio; INDETERMINADA sobre impacto
Limit: The ministry describes an introduced process, not measured outcomes, audited tenders or demonstrated prevention of vendor lock-in.
Permitted interpretation: A government tries to retain operational control and expertise when adopting external systems, rather than merely buying them.
06 OCT 2026 · H26

The UN Human Rights Council adopts a dedicated AI and human-rights resolution

The United Nations Human Rights Council adopted resolution 63/15 without a vote, addressing AI's implications for promoting and protecting human rights. The resolution treats AI governance as a human-rights question, not merely a technical-performance issue.

Why it matters. The question of what machines should do becomes a question of rights, responsibilities and people affected.

FRONTIERS OF AUTONOMY AND CONTROL

PATIELE AT THAT MOMENT · Patiele continues documenting human music authorship and artificial management.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: United Nations · international human-rights framework
Vector: RIGHTS / GOVERNANCE
Evidence type: UN resolution A/HRC/RES/63/15 and official register
Event date: 2026-10-06
Documentary confidence: ALTA
Limit: A Human Rights Council resolution is not a binding treaty and does not prove implementation in each country.
Permitted interpretation: The question of what machines should do becomes a question of rights, responsibilities and people affected.
22 OCT 2026 · H27

Barclays announces roughly 5,000 job cuts amid AI-enabled restructuring

The British bank announced roughly 5,000 job cuts, about 5% of its workforce, within a cost-reduction programme involving AI and other technologies. The disclosure came with third-quarter results.

Why it matters. AI-enabled business reorganisations already coincide with measurable employment decisions, although the causes must be disentangled.

PATIELE AT THAT MOMENT · The experiment continues exploring work an independent artificial organisation can perform.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: Barclays · UK banking
Vector: EMPLOYMENT / RESTRUCTURING
Evidence type: Company announcements reported by Reuters
Event date: 2026-10-22
Documentary confidence: ALTA
Limit: The 5,000 cuts cannot all be attributed solely to AI: financial, organisational and other technological drivers also matter.
Permitted interpretation: AI-enabled business reorganisations already coincide with measurable employment decisions, although the causes must be disentangled.
23 OCT 2026 · H22

Italy turns AI oversight into concrete authorities and duties

Italy's Legislative Decree 179/2026 entered into force, defining supervisory authorities and sanctions, providing for training and including AI used in work organisation within workplace risk assessment.

Why it matters. Regulation starts specifying who oversees concrete systems and under which responsibilities.

FRONTIERS OF AUTONOMY AND CONTROL

PATIELE AT THAT MOMENT · Patiele continues developing its artificial organisation under human decisions.

TECHNICAL RECORD · LIMITS · SOURCES
Vector: REGULATION / LIMITS · WORK
Documentary confidence: HIGH for the observed event
Limit: Entry into force does not establish uniform enforcement or measure actual changes in workplace practice.
Permitted reading: Regulation starts specifying who oversees concrete systems and under which responsibilities.
28 OCT 2026 · H28

New York restricts how AI companion chatbots may behave toward minors

New York signed bill S9051B restricting unsafe features of AI companion chatbots used by minors, including misleading human-like presentation, emotionally manipulative interactions and certain harmful responses. Application is scheduled for 1 June 2027.

Why it matters. Conversational systems are increasingly regulated as relationships that can affect vulnerable people, not just information services.

FRONTIERS OF AUTONOMY AND CONTROL

PATIELE AT THAT MOMENT · Affective human–AI interaction remains an external debate, distinct from IRIS's operations.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: New York State · under-18 users
Vector: CHILDREN / RELATIONSHIPS
Evidence type: New York legislative record and sponsor's official statement
Event date: 2026-10-28
Documentary confidence: ALTA
Limit: Signing is not evidence of compliance or efficacy; enforcement and interpretation remain future questions.
Permitted interpretation: Conversational systems are increasingly regulated as relationships that can affect vulnerable people, not just information services.
29 OCT 2026 · H29

All 53 WHO European Region member states adopt a roadmap for AI in health

Governments across WHO's European Region unanimously adopted its first regional roadmap for AI in health, covering 2026–2033 and prioritising institutional responsibility, ethical governance, data infrastructure, training and evaluation.

Why it matters. Medical AI becomes a matter of intergovernmental commitments about conditions of use, not just product innovation.

FRONTIERS OF AUTONOMY AND CONTROL

PATIELE AT THAT MOMENT · Patiele documents a period of growing debate about human oversight.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: 53 WHO European Region member states · health systems
Vector: HEALTH / INSTITUTIONS
Evidence type: WHO/Europe release and regional roadmap
Event date: 2026-10-29
Documentary confidence: ALTA
Limit: A roadmap is not evidence of deployed systems or clinical improvements. The WHO European Region includes Central Asian countries and is not the EU.
Permitted interpretation: Medical AI becomes a matter of intergovernmental commitments about conditions of use, not just product innovation.
29 OCT 2026 · H30

AI-assisted US political ads surpass $187 million in documented spending

The Wesleyan Media Project identified at least 289 US political ads generated or enhanced with AI and more than $187 million in spending during the 2026 election cycle, using visual review, media reports and AI disclosures.

Why it matters. AI use becomes a material and measurable practice in election communication, not merely a hypothetical threat.

PATIELE AT THAT MOMENT · Patiele continues documenting authorship, mediation and public trust.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: US election campaigns · 2026 cycle
Vector: DEMOCRACY / CAMPAIGNS
Evidence type: Wesleyan Media Project ad-level tracking and spending data
Event date: 2026-10-29
Documentary confidence: ALTA sobre el mínimo identificado
Limit: The dataset undercounts some uses and combines enhancement, generation and deepfakes. Spending does not prove persuasion or determine the 3 November election outcome.
Permitted interpretation: AI use becomes a material and measurable practice in election communication, not merely a hypothetical threat.
31 OCT 2026 · H31

Oakland public workers secure tentative AI-use protections through bargaining

Oakland announced a tentative three-year agreement with IFPTE Local 21 including AI-use protections, following a strike authorisation vote and before a planned 5 November walkout. The union withdrew its strike notice.

Why it matters. Workers are not only reacting to completed job displacement; they negotiate AI conditions and safeguards in advance.

FRONTIERS OF AUTONOMY AND CONTROL

PATIELE AT THAT MOMENT · Patiele's organisation also raises questions about permissions and human work.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: Oakland city government · unionised municipal employees
Vector: WORK / BARGAINING
Evidence type: Official city announcement and independent labour reporting
Event date: 2026-10-31
Documentary confidence: ALTA sobre el acuerdo provisional
Limit: The city release describes a tentative deal pending union ratification and City Council approval. It does not prove workers have already been retrained or protected in practice.
Permitted interpretation: Workers are not only reacting to completed job displacement; they negotiate AI conditions and safeguards in advance.
31 OCT 2026 · H38

East African ministers adopt regional AI strategy and Kiswahili-related higher-education initiatives

East African Community ministers adopted a regional AI strategy and an AI foundation module for higher education, and directed the development of certification benchmarks for Kiswahili-language AI systems during a sectoral council meeting in Kampala.

Why it matters. Social adaptation is not only using imported models: it also involves educational capacity and local benchmarks for a regional language.

PATIELE AT THAT MOMENT · Patiele approaches the late-autumn phase of its artistic cycle.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: East African Community · higher education and languages
Vector: LANGUAGE / EDUCATION / AUTONOMY
Evidence type: Primary East African Community ministerial communiqué
Event date: 2026-10-31
Documentary confidence: ALTA
Limit: Approved strategies and modules do not prove rollout to every university or completed certification schemes. This does not represent all of Africa.
Permitted interpretation: Social adaptation is not only using imported models: it also involves educational capacity and local benchmarks for a regional language.
06–07 NOV 2026 · H23 DISCOVERED LATER

Europe-wide protests challenge large data-centre expansion

Coordinated demonstrations across several European countries challenged large new data-centre projects over their use of water, electricity and land, with actions in Amsterdam, Zaragoza and other cities.

Why it matters. The expansion of computing stops being abstract and becomes contested on specific territories.

PATIELE AT THAT MOMENT · Patiele continues while AI infrastructure becomes a civic controversy.

TECHNICAL RECORD · LIMITS · SOURCES
Vector: INFRASTRUCTURE / TERRITORY · RESISTANCE
Documentary confidence: HIGH for the observed event
Limit: The protests do not represent a unanimous European position; not all data centres serve AI exclusively, and they do not prove widespread cancellations.
Permitted reading: The expansion of computing stops being abstract and becomes contested on specific territories.
16 NOV 2026 · H32

Reuters journalists authorise a possible strike in a dispute that includes AI safeguards

Reuters Guild journalists voted on 16 November to authorise a possible strike and published the result on 17 November: almost 96% participated, and 94% of voters supported authorisation. Demands include human oversight of AI content, compensation for displacement and a share of licensing revenue from journalists' work.

Why it matters. Journalists are bargaining not just over whether AI is used, but who controls it and who shares in the value of human work.

FRONTIERS OF AUTONOMY AND CONTROL

PATIELE AT THAT MOMENT · Patiele prepares public presentations on the division of labour between people and artificial systems.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: Reuters Guild · approximately 300 media workers
Vector: PRESS / WORK
Evidence type: NewsGuild statement and independent press coverage
Event date: 2026-11-16 · Observed: 2026-11-17
Documentary confidence: ALTA
Limit: Strike authorisation is not a walkout. The source records the union's position in bargaining that also concerns pay and other conditions.
Permitted interpretation: Journalists are bargaining not just over whether AI is used, but who controls it and who shares in the value of human work.
19 NOV 2026 · H33

EU education ministers approve conclusions on learning, assessment and AI literacy

The Council of the EU adopted conclusions calling for critical AI literacy across the lifespan, teacher training, adaptation of curricula and assessment, workforce preparation and safeguards for human agency and equity.

Why it matters. Education authorities acknowledge they must rethink what they teach and how they assess learning.

PATIELE AT THAT MOMENT · The musical cycle and artificial organisation remain a real case of independent learning and practice.

OPEN RECORD · EVIDENCE · LIMITS · SOURCES
Scope: EU Council · education and lifelong training
Vector: EDUCATION / LEARNING
Evidence type: Adopted EU Council conclusions and official release
Event date: 2026-11-19
Documentary confidence: ALTA
Limit: Council conclusions are non-binding political guidance, not a compulsory EU-wide curriculum or proof that schools have already changed examinations.
Permitted interpretation: Education authorities acknowledge they must rethink what they teach and how they assess learning.
MAY 2026 · PRE-PATIELE BASELINE

The close relationship with chatbots already existed before the cycle began.

A European survey indicated significant use of chatbots for personal and emotional support among young people. This serves as a baseline, not as evidence of dependence or harm.

May 25, 2026 · Work / India

University degrees are starting to compete with direct demonstrations of AI skills.

In global business hubs in India, a report cited by Reuters found that 40% of employers preferred demonstrable AI skills or certifications over a degree, and another 32% gave them equal weight. Executives also described pressure on entry-level positions. This is a sectoral and geographical trend, not a universal labor market law.

JUN 2026 · AMBIVALENCE

Enthusiasm and nervousness can coexist in the same people

International data from Ipsos makes it impossible to reduce human reaction to a simple acceptance or rejection curve.

June 12, 2026 · Proposal / Canada

Chatbots are beginning to be discussed as systems with obligations to intervene in human crises.

Following intense public pressure, Canada introduced legislation that would create a digital regulator and require chatbots to reduce the risk of harmful content and incorporate intervention measures for conversations about suicide or self-harm. The bill was criticized for its shortcomings and timelines; it remains as an institutional signal, not as current law.

July 23, 2026 · WORK / COUNTERWEIGHT

The transformation of the labor market does not yet equate to an overall collapse in employment.

Adecco argued, citing OECD data, that there was no evidence of an overall "job apocalypse" and cautioned that some companies might invoke AI to justify cuts also due to performance issues or restructuring. This serves as a safeguard against reinterpreting 2026 based on future projections.

AUG 2026 · WORK

Signs of replacement do not yet equate to proven mass unemployment

A British survey found a reduction in entry-level jobs attributed in part to AI and automation. The economic cycle and other factors also play a role.

September 10, 2026 · Productivity

Automated productivity does not always immediately translate into layoffs

Wipro reported released capacity equivalent to about 20,000 people and spoke of redeployment, noting that automation, replacement and distribution are different variables.

September 15, 2026 · Bottom-up Adoption

Some working people are starting to buy their "enhanced capabilities" with their own money.

A Deloitte survey estimated that one in six British workers personally paid for AI tools to use at work, with aggregate spending approaching £1 billion annually. It's a national survey, not a global measure, but it shows adoption that doesn't necessarily come from employers.

17 SEP 2026 · PROPOSAL / EU

The “AI companion” is beginning to emerge as a specific regulatory category for minors

The European Commission proposes that AI companions and chatbots aimed at children under 18 be subject to specific rules and remain disabled by default. The Kids Act must be negotiated with the Member States and the European Parliament; it is filed as a regulatory guideline, not as current law.

13 OCT 2026 · CONTEXT

Deezer announces a forthcoming filter for fully AI-generated music

The search filter was announced for launch before the end of 2026, not verified as already live in October. Deezer already excluded detected fully AI-generated tracks from recommendations; this does not imply banning partly AI-assisted human music.

28 OCT 2026 · CONTEXT

Some teachers adopt oral exams and in-class writing as chatbots spread

Reuters documented specific US teachers who changed how they assess student reasoning. This is not a representative census or a nationwide education reform.

2026-06 · PUBLIC SECTOR / BASELINE

OECD records AI use somewhere in 35 of 36 national governments surveyed

The OECD's 2026 Digital Government Outlook reports AI use in at least one government domain in 35 of 36 countries examined. Much of the underlying observation concerns 2025: a baseline, not a new change caused during Patiele.

29 OCT 2026 · HEALTH / SURVEY

US survey shows qualified support for medical AI and preference for human decisions

Pew Research Center published a survey fielded 22–28 June 2026. Respondents showed conditional support for medical AI and reservations about oversight by companies and governments. These are US attitudes, not clinical results or a global measure.

20 NOV 2026 · TERRITORY / LOCAL CONSULTATION

Nearly 600 submissions challenge a proposed Australian AI data-centre site

Reuters reported nearly 600 public submissions on Queensland's proposed Western Downs Digital Park, with Anthropic an anchor tenant: two supported it, one was neutral and the rest objected or sought delays and safeguards. This is not an Australian referendum nor evidence the project was cancelled.

22 JAN 2026 · BASELINE / EAST ASIA

South Korea enacted its AI Basic Act before Patiele began

South Korea's AI Basic Act took effect on 22 January 2026, four months before Patiele began, introducing transparency and oversight duties with a grace period for some sanctions. It is a baseline, not a change caused during the project's cycle.

22–24 SEP 2026 · CONTEXT / INFORMATION PROVENANCE

Dark Sourcery: fraudulent customer-service numbers enter AI-generated answers

Vigilance Security researcher Ariel Simon documented web pages, posts and documents presenting fake support numbers and login links as official contact information for airlines, banks and other businesses. During some tests, assistants repeated the deceptive information when answering legitimate queries. Results varied, and the team described fraudulent calls and limited reports of harm without a reproducible victim count.

Why this matters: AI-assisted information retrieval can lend apparent credibility to manipulated sources, even when the models themselves were not breached or instructed to disobey safeguards.

Limit / counterevidence: The 'at least 374 brands' figure is the research team's own count, not an independent audit and not 374 confirmed enterprise breaches. Total victims and losses have not been established; the widely repeated 91/92% verification statistic is not reused without its primary methodology. Answers were not deterministic across repeated queries.

Classification: CONTEXT / CONTROL — no new main event

Same clock, different field observed.
This page does not attempt to explain what machines do. For that, there is its sister page: State of Machines in the Outside World →

Data and provenance. Download the structured corpus (JSON) ↗

How to read this file

Unit: an observable and dated human change: law or rule, litigation, strike, protest, reform, institutional reorganization, infrastructure, or materially modified cultural boundary.

Historical criterion: we look for events that can be read retrospectively as "this was already changing." They don't have to be the biggest news story of the day; they just have to leave a verifiable mark on how people or institutions acted.

We do not: turn an individual opinion into "what humanity thinks", read silence as rejection, treat proposals as current laws, or impose a simple story of fascination → fear → regulation.

Contradictions: they are part of the record. Adoption and resistance, automation and employment, integration and prohibition can coexist on the same date.

Relationship with Patiele: the time rail only allows us to know what was happening within the experiment at the same time. It does not imply causality.

Same time.