Does Deloitte's advice on AI governance just happen to point toward... hiring Deloitte?
How does Deloitte's commercial interest shape its framing of governance urgency?
This asks how a consultancy that sells AI governance and transformation services, like Deloitte, might frame the urgency of AI governance in ways that also serve its business. The corpus has no notes on Deloitte itself, so this answer draws on close parallels: how other interested parties frame regulation, and how 'urgency' gets built rhetorically.
This asks how a firm that sells AI advisory services, like Deloitte, might frame the urgency of AI governance in ways that also serve its business. The collection has nothing on Deloitte's reports or positioning, so it can't answer the question directly. What it does have is a sharp parallel case, plus some tools for reading any governance pitch from a party with something to gain.
The closest parallel is Can industry self-regulation slow AI without government enforcement?. Karpf looks at Anthropic's proposal for industry-led 'pacing' of AI development and makes two points that carry over to consultancies. First, the governance design being proposed tends to favor whoever proposes it. Second, oversight by embedded evaluators, modeled on banking supervisors, only works because regulators can impose fines. Without that enforcement, it is just advisory work. Applied to a consultancy, the question becomes: does the urgency point toward government rules with penalties, or toward 'readiness' and self-assessment, which are things clients buy from consultants? Those two lead to very different kinds of governance.
A second way in is to treat urgency framing as rhetoric. How do logos, ethos, and pathos shape AI explanations? argues that any explanation of AI works on three channels at once: logic (the evidence), credibility (who is speaking), and emotion (fear of falling behind, fear of exposure). A consultancy's governance white paper can be read the same way. Ask which channel carries most of the weight. If the logical case is thin and the emotional and authority appeals are heavy, the urgency may have more to do with sales than with the risk itself. Do platforms inevitably decline through value extraction cycles? gives a related idea: a lifecycle in which value shifts from users toward whoever controls the intermediary layer. Advisory firms that set themselves up as gatekeepers of 'responsible AI' sit in a similar intermediary position.
The less obvious finding is that the strategic advice consultancies sell is itself being reshaped by AI, and not always for the better. Do LLMs consistently favor the same strategic choices regardless of context? found that LLMs recommend the same strategic side almost regardless of context, reusing trend-coded vocabulary instead of analyzing the situation. Do newer frontier LLMs actually make better strategic decisions? found that newer models favor short-term profit over long-term investment. Does AI enter strategy where reasoning is deepest or most measurable? argues that AI gains trust in strategy work where its results are easy to measure, not where its reasoning is deepest. Put together, these suggest that governance urgency framed in fashionable, easily measured terms may say less about the risk landscape than about what can be packaged and delivered.
In short, the collection gives you ways to read a Deloitte governance report critically, but no analysis of one. Treat this as a gap in the collection, not a settled answer.
Sources 6 notes
Karpf argues that Anthropic's pacing proposal benefits the company proposing it and that embedded evaluators, modeled on banking supervisors, fail without state enforcement backing them—analogous to how banking oversight works only because regulators can impose fines.
Aristotle's three appeals map onto explanation design across two goals (how AI works, why AI merits use), creating a 3×2 space where every explanation loads all three channels simultaneously. Naming these rhetorical channels lets designers account for unintended persuasive effects.
Doctorow identifies a three-phase lifecycle where platforms initially benefit users, then exploit business customers, then extract shareholder value. Amazon Marketplace, Facebook, and Twitter exemplify the pattern, though the research provides illustrative rather than sampled evidence.
Across 15,000 simulations, six LLMs recommended the same strategic choice in every tension tested. Industry context shifted bias only 11%, while option order—a framing artifact—shifted results 19%, revealing that models recombine trend-coded vocabulary rather than analyze context.
Mid-to-late 2025 frontier models scored below earlier models and MBA students on a strategy simulation, systematically favoring immediate profit extraction over uncertain future bets.
Show all 6 sources
Csaszar et al. argue a dual-ladder framework shows AI gains strategic discretion based on measurable performance at predictive levels, not causal depth. The causal and delegation ladders move separately: AI becomes trusted where forecasting suffices, regardless of whether it achieves true causal reasoning.
Papers this line draws on 8
The research behind the notes this line reads — ranked by how closely each paper relates.
- How Well Can AI Do Strategy? Empirical Benchmarking Using Strategy Simulations
- When Artificial Intelligence Does Strategy: Learning, Good Times, Lock-in, and Human-Driven Strategic Renewal
- AI-Augmented Strategic Decision-Making Under Time Constraints: An Experimental Study on Mental Representations and Strategic Foresight
- The Strategic Foresight of LLMs: Evidence from a Fully Prospective Venture Tournament
- Can AI Do Strategy?
- Rhetorical XAI: Explaining AI’s Benefits as well as its Use via Rhetorical Design
- Your AI Strategy Advisor Is Giving Everyone the Same Advice
- AI Sycophancy and Decisions