Decisions we actually work with — each with its source and a short note on what matters in it. Older entries are kept in the archive.
The German Distance Learning Protection Act is set to be repealed: the draft bill provides for entry into force on 1 July 2027 and full expiry at the end of 30 June 2028 — the draft does not change the law as it currently stands.
General consumer protection under the Civil Code is to take the place of the FernUSG; a certification procedure is envisaged for existing accreditations (§ 27 FernUSG-E). The published text contains no express provision for contracts already concluded.
until any entry into force, § 7(1) FernUSG and the case law of the Federal Court of Justice continue to apply; limitation periods remain governed by the general rules. Analysis in German.
Anyone advertising an "up to" discount must present the essential basis of calculation so that consumers can realistically assess their own benefit — a note in the small print at the end of the advertisement may not be sufficient.
A mail-order pharmacy advertised an "E-Rezept-Rabatt. Automatisch bis zu 20 € Zuzahlungen sparen" ("e-prescription discount — automatically save up to €20 in co-payments"), whereas the discount was in fact tiered by medicine price (€2.50 up to €59.99, €5 up to €249.99, €15 up to €999.99 — the advertised €20 only above €1,000). The Frankfurt Higher Regional Court treated this as withholding material information within the meaning of § 5a UWG (German unfair competition law): the tier table appeared only in small type at the end of the advertising sequence, and purely audible versions did not convey it at all — the ruling was issued in expedited proceedings and is not open to appeal.
Anyone advertising maximum amounts or "up to" figures should place the conditions where the eye-catching claim is actually perceived — the wider the gap between the headline maximum and the realistic standard case, the higher the risk of a warning letter.
Providing test seals such as "FOCUS Top Physician" for a licence fee is a commercial practice under German unfair competition law — and where health is concerned, particularly strict standards of accuracy, clarity and unambiguity apply to both the testing procedure and the seal's design.
The Wettbewerbszentrale (competition watchdog) had challenged the paid licensing of the "FOCUS Top Mediziner" and "FOCUS Empfehlung" seals; the Federal Court of Justice set aside the appeal judgment that had dismissed the claim and remanded the case to the Munich appeal court. Publishing the doctor rankings themselves remains protected press activity — but marketing the seals as advertising tools is assessed separately and can be impermissibly misleading if the testing procedure does not support the quality claim or the seal conceals its limitations (such as reliance on self-reporting and peer recommendations).
Anyone awarding seals or rankings, or advertising with licensed badges, should review and document the testing methodology and seal design — the stronger the health connection, the stricter the standard.
Whether an online coaching programme is "distance learning" depends on the balance of what the contract owes: where synchronous live and in-person elements clearly outweigh video material, the FernUSG does not apply — and a participant who claims self-study must prove it.
In the case decided, roughly 53 hours of synchronous or in-person teaching stood against about 31 hours of on-demand video; a compendium without explanation was not treated as knowledge transfer. The participant's claim failed, the ruling is final — and the senate expressly abandons its earlier line (6 U 46/24).
What matters is what the contract owes, not how the programme was used — the contract text, the programme overview and the learning-platform access are the decisive evidence; providers relying mainly on recorded content remain exposed (BGH III ZR 137/25).
The confirmation page of an online cancellation may contain nothing beyond the form and the confirmation button — no references to cancellation alternatives such as a contract pause; otherwise it violates § 312k BGB.
The case concerned the online cancellation flow of a fitness-studio operator: after clicking the cancellation button, the confirmation page displayed not only the cancellation form but also alternatives such as a fee-free contract pause. The Federal Court of Justice clarified that the confirmation page serves exclusively to collect the required details and submit the cancellation declaration — additional content capable of deterring consumers from their decision to cancel is incompatible with § 312k BGB.
Anyone offering consumers paid continuing obligations online (subscriptions, memberships, SaaS) should review the entire cancellation flow for extra content — retention offers do not belong on the confirmation page, and violations can be pursued under unfair-competition law.
Misleading statements by an AI chatbot are attributed to the company — the chatbot is not a "third party", and clean training data does not exonerate.
The OLG treated the misleading chatbot statements (fabricated medical specialist titles) as an unfair commercial practice under § 5(1), (2) no. 3 UWG and attributed them to the operator — careful selection of training data is irrelevant, since hallucination is inherent in the generative process. An appeal on points of law (Revision) to the BGH has been admitted; the judgment is not final and binding.
Anyone operating a generative chatbot with external effect should ground its outputs in a verified data set and block sensitive statements (titles, prices, availabilities) — a disclaimer does not replace attribution.
Expulsion clauses against manager-shareholders are valid where the shareholding is tied solely to the corporate office.
The BGH adheres to the basic rule that expulsion without compensation is contrary to public policy (§ 138 BGB), but permits call options on pure management shareholdings that carry no independent weight. Pricing and compensation questions do not affect the validity of the clause but are assessed separately.
Leaver and call-option clauses in PE/VC and management-participation agreements can be drafted with legal certainty where the shareholding is cleanly coupled to the corporate office.
Synchronous live online coaching does not fall under the Distance Learning Protection Act (FernUSG).
What is decisive is the content of the contract, not how it is actually carried out. Live calls with a genuine opportunity for interaction remain valid — programs that are predominantly pre-recorded, by contrast, are void.
A reclaim is worthwhile only after a precise analysis of the format; a blanket "contract void" no longer holds.
AI training datasets are permissible under the TDM exceptions — an opt-out formulated only in natural language was not "machine-readable" in 2021.
The OLG confirmed the permissibility of dataset creation (§§ 44b, 60d UrhG) but denied that the reservation of use was machine-readable for 2021, leaving open whether natural language would suffice today. An appeal on points of law (Revision) to the BGH has been admitted and filed.
Rightholders should set opt-out reservations in technically machine-readable form; for AI providers, the legal position on training remains in flux until the BGH decision.
A managing director's tort liability (§ 826 BGB) for a fraudulent scheme does not end with removal from office — it also covers contracts concluded only afterwards.
The Second Civil Senate held a departed managing director liable for an investment contract signed only after his removal, because he had built up the harmful scheme during his term of office and continued to play a key role in it afterwards (§ 826 BGB). What matters is not the formal period of appointment but the continuing responsibility for the danger once created.
Directors stepping down should document and cut off liability exposure from processes already set in motion — and review D&O "tail coverage", because removal from office alone offers no protection.
Where a language model outputs memorized song lyrics almost verbatim, the operator is liable — not the user.
The court treated the output of memorized texts as reproduction (§ 16 UrhG) and making available to the public (§ 19a UrhG); the TDM exception (§ 44b UrhG) does not apply because the content is stored permanently. Responsibility lies with the model operator. An appeal has been announced; not final and binding.
Anyone deploying generative AI should contractually clarify licensing and liability questions regarding training and output content with the provider — responsibility cannot simply be shifted onto users across the board.
The FernUSG also protects businesses — coaching contracts without ZFU authorization are void.
Structured online mentoring programs with monitoring of learning success require authorization. Where authorization is lacking, the contract is void from the outset under § 7 FernUSG and amounts paid are fully recoverable.
B2B customers, too, can invoke this protective statute.
Even the mere loss of control over data gives rise to non-material damages.
Concrete misuse of the data is not required; the BGH considers around EUR 100 per affected person to be reasonable.
Data protection incidents scale into substantial risks through mass claims — sound data management and incident processes become a liability question.
An AI cannot be named as an inventor — but AI-assisted inventions are patentable where a natural person is named as the inventor.
An inventor within the meaning of § 37 PatG can only be a human being; the use of an AI to find the technical teaching does not preclude patentability, as long as there is a substantial human contribution.
In AI-assisted research and development, document the human inventive contribution and name it correctly in the application.
Where the employer permits voluntary ChatGPT use via private accounts, there is no co-determination right — its introduction as a workplace tool, by contrast, does trigger one.
As the employer had no access to the data accruing at the provider, the court denied a co-determination right under § 87(1) no. 6 BetrVG; the employer's introduction of AI as a technical device, by contrast, is subject to co-determination.
Anyone rolling out AI tools across the business should involve the works council and govern usage in an AI policy or works agreement.
Automated scoring is a "decision in the individual case" under Art. 22 GDPR.
The mere calculation of a score value already falls under the prohibition on automated individual decisions where third parties decisively rely on it.
Anyone deploying AI-assisted assessment systems needs a legal basis, transparency about the logic involved, and a right of intervention for the data subject — a mandatory topic for every AI implementation.