AI-washing allegations cost little to avoid

On September 29, Bloomberg Law, Business & Practice published a report titled Steps for Avoiding AI-Washing Allegations Cost Little, Save Much. The report identifies avoiding allegations as a low-cost exercise, but its published title does not identify a company, a jurisdiction, a specific allegation, or the steps involved. That makes the immediate news narrower than the headline may suggest: Bloomberg Law, Business & Practice has put the cost of avoiding AI-washing under scrutiny, not established that any particular firm has avoided liability or reputational damage.

That matters because AI-native legal work is viable only when the work can survive verification and accountability. A low-cost compliance step can still carry a high-value consequence if it prevents inaccurate legal information from reaching a client or a court. The Solicitors Regulation Authority says it has identified cases involving misuse of AI, particularly inaccurate information and client confidentiality, while also saying many solicitors use AI safely and responsibly. The point is not that AI-washing itself causes a sanction. It is that cheap precautions do not make bad legal work cheap: the cost of an error remains attached to the work and to the person or organisation responsible for it.

The report establishes a warning about the economics of avoiding AI-washing allegations, and nothing more. It does not establish what conduct triggers an allegation, what a compliant disclosure looks like, how much the relevant steps cost, or whether those steps prevent inaccurate output. The Solicitors Regulation Authority’s account establishes that misuse cases exist and identifies inaccurate information and client confidentiality as concerns; it does not connect those cases to AI-washing or to any AI-native law firm. A stronger reading would require a named allegation, a filing, an order, or a measured comparison between the cost of verification and the cost of an error. None is identified here. AI does not eliminate the consequences of bad legal work, but this event does not quantify those consequences.

The next fact to watch is a named allegation or filing that identifies the AI statement at issue, the verification or disclosure step that was missing, and the resulting cost. A regulator’s guidance would sharpen the standard; a court order would show how inaccurate AI-generated work is treated after it reaches litigation. The Law Society reports that complaints made with AI are increasing across public services and adding pressure to courts, councils and ombuds services. That pressure is consistent with the broader position that accountability, not merely cheaper production, determines whether AI-native legal work holds up. We should treat the Bloomberg Law, Business & Practice finding as confirmed only when a concrete allegation or order shows what avoiding AI-washing costs and what failing to do so costs instead.