Policy tracker

Assuring quality learning in a gen AI-integrated future: The role of adaptive capabilities

TEQSA (Tertiary Education Quality and Standards Agency)

• Third resource in TEQSA's assessment reform series (24 June 2026), shifting emphasis from securing assessment against gen AI to developing student capabilities.
• Names evaluative judgement, critical thinking and ethical reasoning as adaptive capabilities tertiary education should deliberately develop.
• Offers evidence-informed approaches for assuring learning quality where gen AI is embedded in study and future work, under the Threshold Standards.

Code of Practice on marking and labelling AI-generated content

European Commission · European Commission

• The European Commission published a voluntary Code of Practice on marking and labelling AI-generated content on 10 June 2026.
• It helps providers and deployers meet AI Act Article 50 transparency duties, which apply from 2 August 2026.
• It covers machine-readable marking by providers and the disclosure or labelling of deepfakes and public-interest text by deployers.

AI Adoption Plan: Creative Industries

UK Department for Science, Innovation and Technology · AI Champion for the Creative Industries

• DSIT published an AI Adoption Plan for the Creative Industries on 8 June 2026, authored by the sector AI Champion.
• Its lead principle is augmentation first: AI should support human creativity, not displace creative workers.
• It reports 51% of creative businesses use AI versus 33% across all businesses, and sets out eight practical steps.

Rights, IP & training data

EU ratification of the Council of Europe Framework Convention on Artificial Intelligence

Council of Europe

• The European Union ratified the Council of Europe Framework Convention on AI on 15 May 2026, the first binding international AI treaty.
• The Convention entered into force on 1 November 2025 and covers AI across public and private sectors, with education within its scope.
• Parties must ensure AI lifecycle activities comply with human rights, democracy and the rule of law, including risk assessments and remedies.

Governance & guidance

EU AI Act Omnibus simplification agreement: education compliance deferred to 2 December 2027

European Commission

• EU institutions reached a provisional omnibus simplification agreement for the AI Act on 7 May 2026, pending formal adoption.
• Compliance obligations affecting high-risk education AI systems are deferred to 2 December 2027 under the agreement.
• The package also bans nudification apps as part of citizen-protection measures accompanying the simplification.

Governance & guidance

University of Surrey: AI embedded in every degree from September 2026

University of Surrey

• From September 2026 every University of Surrey degree will include discipline-specific AI teaching, from foundation year to postgraduate level.
• Creative practice is named alongside engineering, law, business and clinical disciplines as areas where core competencies will be preserved.
• Assessment is being redesigned institution-wide around process, decision-making and applied problem-solving rather than outputs.

Teaching, curriculum & assessment Governance & guidance

UK Report on Copyright and Artificial Intelligence: government drops opt-out TDM exception

UK Department for Science, Innovation and Technology (DSIT)

• The UK government formally dropped its preferred opt-out text and data mining exception in its statutory report on copyright and AI.
• The decision followed 11,520 consultation responses; the government will monitor industry-led licensing arrangements instead.
• No immediate legislative change was enacted; the report and impact assessment set the evidence base for future steps.

Rights, IP & training data Governance & guidance

European Parliament resolution on copyright and generative AI (TA-10-2026-0066)

European Parliament

• The European Parliament adopted its resolution on copyright and generative AI on 10 March 2026 by 460 votes to 71.
• It calls for full transparency on copyright-protected works used in AI training, including an itemised list of each work.
• Parliament asks the Commission to examine fair and proportionate remuneration of rights holders for past unlicensed training uses.

Rights, IP & training data Governance & guidance

Screen Australia AI Transparency Statement (stakeholder provisions)

Screen Australia

• Screen Australia requires any stakeholder use of AI, including by funding applicants, to be declared and consistent with its Terms of Trade.
• Since 2025 all agency funding application forms collect information on AI usage in applications and funded projects.
• The statement rests on Screen Australia's September 2024 AI Guiding Principles covering transparency, ethical use and accountability.

Production & craft Rights, IP & training data Governance & guidance

House of Lords report: AI, copyright and the creative industries

UK Parliament, Communications and Digital Committee

• The Lords Communications and Digital Committee finds UK creative industries face a clear and present danger from unremunerated AI training on copyrighted work.
• It recommends ruling out an opt-out text and data mining exception and making training-data transparency a statutory obligation.
• It urges a fair licensing market, protections against unauthorised digital replicas and style imitation, and sovereign AI capability.

Rights, IP & training data Labour & workforce

Thaler v. Perlmutter: Supreme Court denies certiorari, human-authorship requirement upheld

US Supreme Court / SCOTUSblog

• The US Supreme Court denied certiorari in Thaler v. Perlmutter on 2 March 2026, leaving the D.C. Circuit human-authorship ruling intact.
• The D.C. Circuit ruling stands: US copyright law requires a human author; an AI system cannot hold authorship.
• The ruling is limited to works with no asserted human authorship; human contributions to AI-assisted works are assessed case by case.

Rights, IP & training data

BFI guidance on AI use in funding applications and funded projects

British Film Institute (BFI)

• The BFI does not prohibit AI use in funded work but requires all applicants to be transparent about assistive and generative AI use.
• All BFI funding applications now include questions on AI use in the project and in preparing the application itself.
• Applicants and awardees must warrant that no third-party rights were or will be infringed, extending to any AI use in the project.

Production & craft Rights, IP & training data Governance & guidance

TEQSA regulatory-led framework for AI and assessment

TEQSA (Tertiary Education Quality and Standards Agency)

• TEQSA moves to a regulatory-led approach to generative AI and assessment in Australian higher education from 2026.
• Providers must evidence AI risk management and assessment validity under the Higher Education Threshold Standards.
• The Good Practice Hub resources apply to all disciplines, including studio-based creative-arts programs.

Teaching, curriculum & assessment Governance & guidance

Australian Framework for Artificial Intelligence in Higher Education

Australian Centre for Student Equity and Success (ACSES)

• Seven principles for ethical, equitable and effective AI across Australian higher education.
• Explicitly includes Indigenous knowledges as a core principle for institutional AI adoption.
• Developed by a multi-university team as the sector-wide companion to TEQSA's assessment guidance.

Teaching, curriculum & assessment Governance & guidance

Getty Images (US) Inc v. Stability AI Ltd: UK High Court Judgment

Courts and Tribunals Judiciary (England and Wales)

• First UK ruling on AI training and copyright: no primary infringement where training occurred outside UK territory.
• Court accepted that diffusion models do not store reproductions of training images.
• Limited trade mark infringement found where earlier Stable Diffusion versions reproduced Getty watermarks.

Generative techniques Rights, IP & training data

Japan anime and manga industry joint statement on generative AI

Japan Cartoonists Association and Association of Japanese Animations (English coverage: Animation Magazine / Anime News Network)

• Eighteen anime and manga organisations declared that AI operators must obtain rights-holder permission at training and generation stages.
• The statement explicitly criticised Sora 2 for enabling Ghibli-style generation without permission.
• It is Japan's industry-level counterpart to the US union AI positions the repository already holds.

Rights, IP & training data Labour & workforce Governance & guidance

SAG-AFTRA 2025 Interactive Media Agreement: AI Provisions

SAG-AFTRA

• Separate written and specific consent is required before any digital replica of a performer is created.
• Consent and compensation obligations trigger when output is objectively identifiable as a specific performer.
• Ratified by 95 per cent of members after an 11-month strike centred on AI protections.

Production & craft Rights, IP & training data Labour & workforce

Disney Enterprises Inc et al. v. Midjourney Inc: Complaint

United States District Court (CourtListener)

• First joint copyright action by major Hollywood studios against an AI image generation company.
• Alleges wilful infringement of animated characters through both training and output generation.
• Seeks statutory damages across more than 150 listed works, with Warner Bros filing a parallel action.

Generative techniques Production & craft Rights, IP & training data

US Copyright Office Report on Copyright and AI, Part 3: Generative AI Training

US Copyright Office

• Part 3 of the US Copyright Office AI report addresses whether training generative AI on copyrighted works constitutes fair use.
• The report concludes that fair use turns on the specific facts of each case; some training uses are likely fair, others are not, with market effects central to the analysis.
• The report rejects the argument that AI training is inherently non-expressive, holding that models generating expressive content cannot categorically claim fair use.

Rights, IP & training data

98th Academy Awards Rules: AI and Human Authorship

Academy of Motion Picture Arts and Sciences

• Generative AI tools neither help nor harm a film's chances of Oscar nomination under the 98th cycle rules.
• Judges weigh the degree to which a human was at the heart of creative authorship when evaluating eligible works.
• The human-authorship standard was recommended by the Academy's Science and Technology Council.

Production & craft Rights, IP & training data