Issue 6
Viewpoint
lawsuit · United States
Unsealed filings: Microsoft and OpenAI executives knew AI threatened publishers
Newly unsealed filings in the New York Times lawsuit show a Microsoft executive calling AI the "largest theft of labor" in history, and the Times says OpenAI staff also recognized the threat to publishers. The filings also raise the possibility of an "accidental cover-up" in identifying content from the Times (September 17–19, 2026).
Who is affected and how
- lawyer
- The parties' internal communications can become a central tool of discovery, so targeted requests for similar documents are warranted on the publisher side.
- corporate lawyer
- AI developers' internal remarks should also be weighed when assessing liability and licensing risk.
tool · United States
OpenAI unveils Astra for Law legal AI platform
On September 17, OpenAI released Astra for Law, a version of GPT-6 Astra tailored to legal work and aimed at large firms and legal technology vendors. Reports say it can search more than 230 million…
lawyerThe tool for legal research and workflows can be tried out, but according to Mobilissimo it was not built as a virtual lawyer, so results still need to be checked.corporate lawyerIn procurement, the decisive factors are legal-grade controls, data handling and comparison with competing legal AI products.compliance specialistPrivacy and confidentiality settings should be reviewed before the new platform is rolled out.compliance · European UnionEuropean Commission enforces AI Act: GPAI obligations binding, transparency rules in effect
The European Commission is enforcing the AI Act, with obligations for general-purpose AI (GPAI) providers now binding and transparency rules in effect. Professional commentary says voluntary safety…
compliance specialistOrganisations should review which provisions apply to them based on their roles, product scope and territorial link.corporate lawyerContractual obligations of US and other third-country companies need to be aligned with the EU's scope of application.lawyerGiven the AI Act's penalty regime, compliance due diligence for clients is timely.lawsuit · United StatesAntitrust lawsuit filed against OpenAI, Anthropic, Google and SpaceXAI
An antitrust complaint has been filed against Anthropic, OpenAI, SpaceXAI and Google, according to Politico over a call to regulate the "pace" of AI development. Bloomberg Law reported on the…
corporate lawyerThe competition-law risk of joint regulatory action and industry agreements needs to be reassessed.compliance specialistIndustry consultations and joint statements warrant a competition-law review.ruling · United StatesAppeals court: Microsoft, OpenAI and GitHub not liable for DMCA claims
A US appeals court held that missing copyright management information in AI-generated code does not violate the DMCA, so Microsoft, OpenAI and GitHub are not liable on that claim. The companies also…
lawyerThe prospects of AI-training claims brought under DMCA Section 1202 need to be reassessed.corporate lawyerDevelopers of AI code generators may be in a more favourable position on the DMCA claim.Hungary
lawPoland's AI systems act takes effectcompliance specialistOrganisations operating on the Polish market must prepare for both the national AI authority and data protection obligations.corporate lawyerRegional differences in member-state implementation rules need to be addressed in contracts.data protectionHungarian labour law and data protection articles examine AI use in the workplace and HRcorporate lawyerInternal policies on employer AI use should be drawn up with data protection and labour law considerations in mind.compliance specialistAI support for HR processes should be introduced alongside a data protection impact assessment and classification under the AI Act.lawyerFor employment clients, it is worth clarifying the transparency of algorithmic decisions and the information given to employees.regulationAI regulation comes up at hearings of candidates for head of Hungary's media authority NMHHcompliance specialistThe AI-related priorities of the media regulator's head may shape domestic content-labelling and platform-oversight practice.corporate lawyerCompanies in media should prepare for an obligation to label synthetic content.professionHungarian AI training for lawyers: Edulex course and Wolters Kluwer Libra AcademylawyerMandatory continuing-education credits can also be earned through an AI-themed course.corporate lawyerIn-house legal teams can build AI competence through organised training.lawSlovak parliament moves AI and European data regulation bill to second readingcompliance specialistOn the Slovak market, organisations should prepare for content-labelling obligations and national implementation.corporate lawyerSlovak legislation affects cross-border contractual compliance.data protectionRomanian experts warn of data protection risks in using AI in the justice systemlawyerClient data should be loaded only into tools whose storage location is known and covered by contract.professionHungarian Bar Association piece addresses when to disclose AI use in text and imageslawyerThe bar guidance may serve as a reference in interpreting labelling obligations and professional expectations on AI use.complianceHungarian article sets out legal background to calls made with automated dialling systemscompliance specialistDomestic rules on automated calls should be checked before dialling systems are introduced.corporate lawyerThe contractual and data protection terms of AI-driven customer service and sales calls should be reviewed.
More key stories of the week
profession · United States
AI-flawed court filings rise despite three years of sanctions
According to Reuters, the number of court filings riddled with AI errors is rising, and judges are increasingly frustrated. One tally puts the number of court decisions involving AI-hallucinated…
lawyerVerification steps must be built into the workflow; alongside sanctions, reputational risk is growing.corporate lawyerChecks on outside firms' AI use can be made a contractual obligation.regulation · United StatesUS Senate AI bill could block risky models and override state law
A US Senate AI bill could block risky models and override state rules, while regulation is deadlocked and pressure on Congress is growing. NBC News summarised the proposals lawmakers are weighing on…
compliance specialistBecause federal preemption is possible, state compliance programmes need to be handled with contingency plans.corporate lawyerThe possibility of blocking risky models creates a new risk in development and operations contracts.lawyerDisputes over the relationship between federal and state AI law will become the core of client advice.litigation · globalUMG and Sony file second lawsuit against Suno over new model
Universal Music Group and Sony Music filed a second lawsuit against Suno over the AI music generator's new model. According to Mixmag, Suno admitted using music from YouTube, and a Munich court found…
lawyerParallel US and European proceedings call for a joint strategy.lawsuit · United StatesTimes, OpenAI and Microsoft ask judge to rule on AI fair use before trial
Publishers, OpenAI and Microsoft are asking the court to decide before trial whether AI training counts as fair use, moving the case into a new phase. The publishers argue that AI companies should…
lawyerThe fair use argument moves to an early, document-based stage, so evidence strategy must be adjusted accordingly.corporate lawyerThe outcome bears on the risk assessment of content-licensing models tied to AI products.data protection · European UnionSpanish data protection authority receives first data breach report linked to an AI agent
Spain's data protection authority, the AEPD, says an AI agent chained together a successful login, vulnerability discovery and access to personal data, and may have modified personal data and…
compliance specialistAI agents' access rights and the incident reporting process should be treated as a separate risk.corporate lawyerContracts for agent-based solutions should clarify the controller and liability relationships for data processing.ruling · United StatesFederal judge lets defamation claim over Google's AI Overviews proceed
A federal court in Illinois has allowed a defamation suit against Google by an author and TV producer to continue, over AI Overviews that allegedly made a false statement about them. According to one…
lawyerPersonality-rights claims over AI outputs may be enforceable, so clients should be prepared for this.corporate lawyerThe liability risk for AI summaries built into search is growing.citizenLegal action can be taken against false AI statements.case · United StatesJudge in OpenAI copyright case stays on after selling Microsoft shares
Reuters reported on Ona Wang, the US magistrate judge overseeing pretrial issues in the copyright lawsuit against OpenAI and Microsoft. According to Reuters, a US judge remains on the case after…
lawyerWhere similar interests exist, it is worth checking a judge's financial conflicts of interest in good time.law · United StatesCalifornia enacts child-safety and AI chatbot laws
Newsom signed California's package of laws on children's online safety and AI chatbots, including "Adam's Law". The privacy and AI measures of the 2026 session, as well as Google's legislative…
compliance specialistChatbot and social media providers must build the new child-related obligations into their operations.corporate lawyerThe liability and terms of service for AI chatbot products must be amended in line with the new laws.ruling · United StatesJudge blocks Montana from enforcing deepfake election ad law
A US judge has stopped Montana from enforcing its law on deepfake election ads against a conservative group. According to MediaPost, the judge found the law likely unconstitutional.
lawyerThe ruling offers a basis for challenging state laws restricting AI-generated content on free speech grounds.compliance specialistThe enforceability of state deepfake rules is uncertain, so campaign compliance must be assessed case by case.compliance · United StatesState AI laws and federal preemption: Texas, Colorado and compliance guides for companies
Texas's TRAIGA, the reset of Colorado's AI law and the question of federal preemption continue to create compliance work for companies. An executive order directed the commerce secretary to assess by…
compliance specialistDiverging state rules require a compliance map covering several states.corporate lawyerCompliance with state AI laws has to be planned together with the risk of federal preemption.lawsuit · United StatesUniversal Music Group sues DistroKid over AI-generated music
On September 15, Universal Music Group filed a lawsuit against DistroKid over "unlawful practices" and for flooding platforms with "AI slop". The complaint also raises copyright issues.
corporate lawyerThe contractual and moderation obligations of music distributors and platforms should be reviewed.case · United StatesState Farm lawyer fined $999.99 over AI hallucinations in Los Angeles lawsuit
A lawyer for State Farm was fined $999.99 over AI hallucinations in a Los Angeles lawsuit. The case was reported on September 14.
lawyerCitations must be checked before filing; representing a large client is no exemption.In brief
caseLawyer fined over AI hallucinations in murder case appeallawyerIn a criminal case, an unchecked AI source raises questions of defence counsel's responsibility as well as sanctions.caseManhattan DA seizes 12 celebrity deepfake websiteslawyerProsecutorial domain seizure offers a new tool when representing victims.compliance specialistA review of hosting and platform-provider risks is warranted.citizenDeepfake victims can also seek redress through law enforcement.toolLegora adds legal research and AI-native citator to challenge Westlaw and LexislawyerCompeting legal research platforms give firms more bargaining power on pricing and service.corporate lawyerIn procurement, a citator function matters for reducing hallucination risk.rulingFlorida appeals court condemns "AI slop" in filingslawyerThe substantive quality and concision of filings can be sanctioned, not only false citations.regulationNewsom executive order calls for independent AI auditors and safety assessorscompliance specialistThe introduction of AI audits and an auditor registry could bring a new external-oversight requirement.corporate lawyerThe contractual obligations of California AI developers may expand because of external assessments.regulationUS judiciary develops new guidance on courts' use of AIregulationDigital Omnibus is the first substantive amendment to the EU AI Actcompliance specialistThe amendment may affect the compliance timeline and priorities, so it should be monitored.corporate lawyerContractual compliance clauses should be reviewed in light of the amendments.regulationUK MPs and peers call for new AI law and regulator to address human rights riskslawyerUncertainty in English law on liability for AI harms should be treated as a client risk.compliance specialistFor activity in the UK, firms should prepare for possible new supervisory requirements.professionIsrael's Bar Association sets strict rules for lawyers using AI agentslawyerThe use of AI agents is subject to Bar rules, so office practice should be aligned with them.rulingDelhi High Court refuses immediate stay in ANI v OpenAI copyright caselawyerThe Indian precedent should be taken into account when advising Asian clients on AI copyright.professionMorgan & Morgan to spend at least $1bn on legal AI over 10 yearslawyerLarge firms' own AI platforms can be a competitive advantage in cost-sensitive practice areas.corporate lawyerLaw firms' technology investment is also a factor in management and fee negotiations.data protectionReport: OpenAI contract workers can read ChatGPT conversations, raising privacy concernscompliance specialistUse of consumer AI tools with client data should be banned or restricted by policy.lawyerBecause of client confidentiality, using public chatbots is risky.citizenSensitive data should not be shared with public AI chatbots.lawCalifornia requires ads to disclose AI-generated performerscompliance specialistIn ad campaigns, labelling of AI performers must be built into the production process.regulationGPAI Code of Practice transparency chapter published as Commission readies guidelinescompliance specialistGPAI providers must build the Code's documentation measures and the guidelines into their compliance process.corporate lawyerSigning the Code carries contractual and liability implications.rulingCourt denies xAI preliminary injunction against Minnesota's AI nudification bancorporate lawyerChallenging state-level content restriction laws may be difficult on this path.compliance specialistAI image-generation services need to prepare to comply with such laws.professionSRA issues warning notice on misuse of AIlawyerThe regulatory warning bears on professional liability and disciplinary risk.corporate lawyerIt is advisable to request contractual declarations on AI use from external law firms.caseOklahoma judge's ruling contained AI hallucinationslawyerWhere a ruling is flawed, checking its citations and considering a remedy is advisable.professionLaw schools and firms respond to AI: Touro requirement, free Harvey access, free Claude traininglawyerFree training opportunities help lawyers learn to use AI safely.professionLegal AI startup Harvey valued at $15.5bn, hires Microsoft research chieflawyerThe strengthening of the market-leading legal AI provider influences firms' technology choices.corporate lawyerThe evaluation methods of legal AI developers also matter for procurement decisions.professionNew AI-native law firms launchlawyerAI-native models open up new firm structures and client segments.toolRelativity renames RelativityOne as Relativity aiR; Checkbox launches First Pass contract reviewlawyerAI-based document review and contract checking speed up first-pass review work, but human review remains.corporate lawyerScreening can be automated in in-house contract management; confidentiality matters when choosing a tool.perX and xAI settle with Apple; antitrust claims against OpenAI continuecorporate lawyerThe partial settlement shows how the structure of the claim and the parties' bargaining positions have changed.lawyerThe claim against OpenAI signals the next phase of antitrust litigation over AI.regulationDispute over AI companies' "pace the frontier" regulation call and lobbyingcompliance specialistThe regulatory direction shaped by lobbying should be tracked through preparatory documents and draft bills.lawIreland's domestic AI legislation takes effect; Italy regulates police biometricscompliance specialistMember-state supplementary rules create separate compliance tasks alongside the AI Act.corporate lawyerItalian liability rules call for contractual risk allocation.regulationDOJ will not regulate AI firms through prosecution, Blanche says; Khan says existing law sufficescorporate lawyerCompliance risk should also be assessed in light of executive liability under existing laws.compliance specialistDifferences in regulators' enforcement priorities affect risk assessment.perUniversal and Warner ask court to end US musicians' Udio and Suno lawsuitperCanada's largest music rights group sues AI music generatorlawyerCanadian law has no fair use doctrine, so arguments need a different framework.toolThomson Reuters launches its own AI model trained on proprietary legal contentcorporate lawyerKnowing the origin and training data of vendors' models is part of the procurement decision.lawyerThe market for legal research tools is also widening toward proprietary models built by traditional providers.complianceNvidia, Palantir and other firms restrict use of advanced AI models over privacy concernscompliance specialistCorporate AI-use policies should regulate access to models and data according to risk.corporate lawyerData-handling and intellectual-property safeguards are key issues in contracts with AI providers.professionDeepfakes and AI prompts as evidence in courtlawyerAuthentication of digital evidence and preservation of prompts should be planned ahead in litigation.caseFederal court asks for help on AI-generated child sexual abuse materiallawyerThe criminal-law question is uncertain because of a legislative gap, so arguments for the defence and the prosecution will be case-specific.professionMore than 110 AI copyright lawsuits filed as AI-generated books flood online marketplacesregulationHong Kong to appoint AI commissioner; South Korean court convicts official over AI imagescompliance specialistNew compliance expectations can be expected for activity in Hong Kong because of the AI commissioner and the deepfake rules.lawyerThe Korean ruling is an example of how AI images that are humiliating but not nude are assessed under criminal law.professionSurveys on legal AI adoption: ILTA tech survey and Texas lawyerslawyerThe surveys offer a benchmark for how colleagues use AI.corporate lawyerMarket adoption and caution can both be considered when rolling out AI internally.regulationEU plans child protection law for social media, games and AI chatbots; concerns over transatlantic AI dividecompliance specialistChild-protection compliance of AI chatbots should be prepared in light of the EU draft.corporate lawyerContractual and compliance burdens for smaller developers may increase.data protectionHungarian press examines data-privacy questions behind the '80s-style AI photo trendcitizenBefore uploading a photo, check the terms of data processing and the usage rights.regulationNew York seeks broader AI regulation before its first AI law takes effectcompliance specialistGiven New York's regulatory plans, compliance programmes should be planned ahead of the effective date.lawNew AI rules in Illinois and Connecticut are about to take effectcompliance specialistAffected AI applications should be reviewed ahead of the effective dates.caseIndiana lawyer hit with $10,000 sanction over false citationslawyerSanction amounts are rising, so citation checking is a matter of risk management.