Issue 13 · Nov 2–8, 2026Agent Age News magazine · Other editions · Method · HU / ENG

The impact of AI on law, the courts and the legal profession, curated weekly

Issue 2

Aug 10–16, 2026 · 45 stories · 7 of them from Hungary
Viewpoint

regulation · European Union

EU digital omnibus in force: high-risk AI rules pushed back to December 2027

The Council of the EU gave final approval to the digital omnibus on AI on 29 June 2026, which according to the sources entered into force on 27 July. The deadline for high-risk systems moves to December 2027, but the package also brings a new prohibition and a new GDPR legal basis for bias detection.

Aug 10, 2026 · 20 sources

Who is affected and how

corporate lawyer
Contract and compliance projects for high-risk AI systems must be rescheduled to the new deadlines.
compliance specialist
The compliance roadmap must be updated, separating postponed obligations from those that are already live.
lawyer
AI Act advice to clients must be based on the new deadlines and the newly introduced prohibition.
compliance · European Union

AI Act transparency and enforcement rules apply from 2 August

From 2 August 2026, the transparency obligations of Article 50 of the AI Act apply, requiring chatbots to disclose that they are not human and AI-generated content to be labelled. The procedural…

compliance specialistCompanies must assess which interfaces require chatbot and AI-content labelling, and document this.corporate lawyerSoftware bought from a US provider does not exempt a company from the AI Act, so supplier contracts must be reviewed.case · United States

Hidden AI instructions in court filings: first documented prompt injection aimed at a US court

In a Connecticut case, the plaintiff is alleged to have placed hidden messages in their filings to sway a court that might be using AI in their favour. The man told Newsweek that he was thereby…

lawyerFilings and attachments should be screened for hidden text if they are read with an AI tool.case · global

AI-hallucinated citations in court filings continue, with sanctions

A US federal court has held that ignorance is no excuse for erroneous AI citations, while a Mississippi judge declined to impose new sanctions. According to our sources, more than 1,000 such cases…

lawyerEvery citation and quotation produced with AI must be checked against the original source before filing, because ignorance is no excuse.ruling · European Union

Munich court rules AI music training and outputs infringe copyright in GEMA v Suno

The Munich I Regional Court ruled largely in favour of GEMA against AI music generator Suno on 31 July 2026. According to the sources, it found infringement even where the training took place in the…

lawyerIn cross-border AI training cases, the questions of EU jurisdiction and applicable law gain a new argument.

Hungary

complianceAI Act transparency rules apply to Hungarian businesses from Augustcorporate lawyerFor Hungarian clients, it needs to be clarified which AI content and chatbots fall under the labelling obligation.compliance specialistMap where the company uses AI and set out the labelling practice in an internal policy.lawHungarian government legislative council to consider draft AI lawcorporate lawyerFollow the text of the draft, because the national implementing rules will shape clients' compliance obligations.compliance specialistChanges in the draft will affect domestic supervisory and reporting processes.professionHungarian Bar Association issues position on lawyers' use of AI speech-to-text toolslawyerBefore using AI dictation and transcription tools, confidentiality and data-handling terms must be checked in line with the Bar's position.regulationHungarian bodies in AI Act enforcement: market surveillance, AI Council, media regulator NMHHcompliance specialistYou need to know which Hungarian authority to notify or approach with questions about live chatbots and HR-related AI.corporate lawyerThe high-risk classification of HR systems should be assessed in light of Hungarian regulatory practice.lawBills proposed to penalise deceptive AI content that manipulates identitylawyerThe new or amended offences will be relevant when representing deepfake victims and the defendants concerned.citizenCriminal-law protection against AI content that forges identity may be strengthened.data protectionHungarian data authority NAIH halted a bank's AI emotion analysis, imposing HUF 250 million finecompliance specialistFor voice-analysis AI used in call centres, NAIH's earlier practice means a serious risk of fines must be factored in.corporate lawyerFor a voice-recording analysis project, the data protection impact assessment and the legal basis must be documented separately.regulationPoland to build AI regulatory sandbox with 35 million zlotycompliance specialistThe sandbox allows AI systems to be tested under regulatory supervision.corporate lawyerClients in the region should watch the sandbox's participation conditions.

More key stories of the week

ruling · Asia

Delhi High Court rules in India's first AI copyright case, ANI v OpenAI

Single judge Amit Bansal of the Delhi High Court delivered a judgment on 24 July 2026 in the case of ANI Media and OpenAI, on how copyright applies to the training and operation of large language…

lawyerThe reasoning of the judgment is worth comparing with EU and US practice.compliance · global

Anthropic watermarks Claude's text and file outputs to comply with EU rules

Anthropic has reportedly been invisibly watermarking Claude's outputs since 2 August to meet EU transparency rules. According to Euronews, the marking applies worldwide.

compliance specialistProvider-side watermarking shows how the machine-readable marking requirement is being met, and it is the point to raise in your own supplier assessments.lawyerText produced with Claude carries a watermark, so provenance checks on documents meant for clients need to take this into account.ruling · Asia

Chinese courts side with workers displaced by AI

Chinese courts have ruled in favour of workers displaced by AI, but workers' fears persist, according to the report. The professional question is how employment law treats AI-driven layoffs.

corporate lawyerIn layoffs linked to AI adoption, employment-law requirements on justification and procedure need particular attention.lawyerChinese rulings can serve as a comparative reference when arguing on the employee's side.profession · European Union

Swedish legal AI startup Legora in talks to raise funds at valuation above $10 billion

The Swedish legal AI startup Legora is reportedly in talks at a valuation of about $10 billion, four months after being valued at $5.6 billion. Vendors point to intensifying competition among law…

lawyerWith legal AI platforms gaining value so quickly, firms must decide fast on adoption.corporate lawyerWhen choosing a vendor, pricing and financial stability should also be examined.profession · United States

New York City Bar issues guidance on AI and attorney-client privilege

A New York City Bar Association task force has published a report on how attorney-client privilege and work product protection apply when generative AI is used. The committee is also calling for a…

lawyerFor AI prompts containing client information, confidentiality risk must be addressed in advance.corporate lawyerAn internal AI-use policy should set out which tools may receive confidential material.lawsuit · United States

Babylon Bee sues New Mexico over AI political satire law

The satirical news site Babylon Bee claims that a New Mexico law violates the US Constitution and has filed suit to stop it. The law affects political satire created with AI.

lawyerThe case may set the standard for free-speech scrutiny of state rules on AI content.case · United States

Suno strikes licensing deal with BMG while preparing new label-backed models

Suno has reached an agreement with BMG, which the article describes as the publisher's most significant step towards accepting AI music so far. BMG has previously also sued Anthropic in a federal…

data protection · European Union

French CNIL note on agentic AI and EDPB guidance on web scraping for generative AI

France's data protection authority, the CNIL, has published a note on agentic AI and data protection. The European Data Protection Board (EDPB) has issued new GDPR guidance on web scraping for…

compliance specialistFor scraped training data and agentic AI, the GDPR legal basis and transparency should be reviewed in light of the new guidance.lawyerThe regulators' guidance provides a reference point for clients' data protection impact assessments.regulation · European Union

Irish High Court issues practice direction on use of generative AI

The Irish High Court has issued a practice direction on the use of generative AI in court. The rule affects lawyers and litigants in Irish proceedings.

lawyerIn filings before the Irish High Court, the use of generative AI must be handled in line with the practice direction.law · United States

California AI Transparency Act takes effect

The California AI Transparency Act has become operative, TechInformed reported. The law sets transparency requirements for AI-generated content.

compliance specialistAI providers serving California users need to check the new obligations.corporate lawyerEU and California transparency rules need to be aligned at the contractual liability points.law · United States

California's Frontier AI Safety Act clears final legislative vote

The California Frontier AI Safety Act has passed the final legislative vote. According to the source, the bill concerns safety rules for the largest AI developers.

compliance specialistFrontier-model developers in California must prepare for the new safety requirements.corporate lawyerUntil signing and entry into force, clients need to prepare contractual risk management.regulation · United States

FTC issues proposed policy statement on AI accuracy and output steering

The FTC has issued a proposed policy statement on the accuracy of AI output and output steering. According to a Consumer Law summary, the topic bears on consumer protection compliance.

compliance specialistAI products' accuracy claims and the steering of outputs need to be reviewed for consumer protection risk.corporate lawyerMarketing and product descriptions need to be checked against the FTC's position.

In brief

rulingMassachusetts court: AI output is not work product without attorney involvementlawyerTo protect materials produced with AI, lawyers must document their own directing role.citizenAnyone using AI without a lawyer should expect the output to go unprotected.caseHessian judges found to have visibly copied ChatGPT sources into rulingsperOtter.AI fails to get class-action privacy lawsuit over meeting transcripts dismissedcompliance specialistThe recording and consent practices of AI note-taking tools need to be reviewed.lawyerFor transcription tools used in client meetings, participants' consent needs to be secured.professionAI chips away at the billable hour as AI-native law firms emergelawyerFee models need rethinking, because clients will expect to share in AI efficiency gains.lawUS states tighten deepfake rules: Oregon, New Jersey, LouisianalawyerLawyers representing victims and defendants need to know state-level offences that differ from one another.citizenBecause protection differs from state to state, it is worth learning the rules where you live.toolHarbor launches legal-focused AI deployment service, Harbor Deploycorporate lawyerUsing an implementation service can speed up putting legal AI tools into production.lawyerFirms should weigh the costs of in-house deployment against an external service.toolGoodwin rolls out AI-enhanced tool for venture financinglawyerIn transactional work, the firm's AI tool may change how financing documentation is prepared.perAI copyright cases: OpenAI sanctions dispute and Disney, Warner fight against trimming of caselawyerProcedural developments offer a pattern for handling evidentiary issues.regulationEU opens three channels for reporting AI Act breachescompliance specialistPrepare for competitors or affected parties to file reports, which means an internal escalation procedure is needed.citizenAnyone who spots a violation can choose from three official channels.regulationUS court and prosecutor rules on AI use in filings, from Maryland to GeorgialawyerIn filings before the Maryland court, AI use must be declared in line with the local standing order.data protectionPhilippines extends data privacy rules to AI systems and warns on deepfakescompliance specialistAI providers operating in the region must apply data processing rules to their AI systems as well.regulationBelgium prepares new product liability rules; US states lack uniform AI health liability rulescorporate lawyerFor software and AI products, product liability risk should be addressed in contractual indemnity clauses.compliance specialistFor healthcare AI, differences in liability rules between states need to be mapped separately.professionColumbia Law School updates AI policy with new restrictionslawState chatbot laws: Hawaii's AI companion rule and the risk of a regulatory patchworkcompliance specialistChatbot disclosure and safety requirements that differ by state need to be mapped.corporate lawyerFor multi-state providers, contractual and policy exceptions need to be managed.perFederal lawsuit alleges Grok was used to create child abuse images from a childhood photolawyerIn victim cases, the AI provider's liability can also be an argument.citizenVictims' legal options may include suing the provider.perApple's $250 million AI iPhone settlement to pay out sooner than expectedcitizenAffected iPhone owners should check their eligibility and the conditions for payment.perAmazon v. Perplexity: where agentic AI stands under the CFAAcorporate lawyerFor products that use web agents, platforms' terms of use and CFAA risk need to be reviewed.lawyerLitigation is shaping the arguments over unauthorised access by agentic AI.rulingUS court developments: first agentic AI appellate ruling, early approval of AI document reviewcorporate lawyerCompanies using agentic AI can use the ruling as a starting point for structuring liability and contracts.lawyerIn e-discovery, courts' acceptance of AI-assisted review can be built upon.caseChina rejects AI-based inventive-step argument in patent prosecutionlawyerFor AI-assisted inventions, the case for an inventive step needs to be carefully substantiated.professionIndian judge calls for judiciary-controlled AI platform, flags privacy riskscaseExperts warn against using AI clone of murdered woman's voice in courtlawyerThe opposing side may object to the use of synthetic voice or images as evidence.

← Issue 1 · Issue 3 →