AI and Advisory Services: The Market Begins to Price In the End of Standardised Work

Artificial intelligence is penetrating the core of professional services. Not because it can already replace a consultant, an advisor or a corporate transformation team in their entirety. The point is simpler and more uncomfortable: a portion of the work that was sold as specialised is becoming fast, replicable and low-cost.

For years, the large consulting firms built value on the ability to gather data, read documents, compare scenarios, prepare presentations, build models and distil all of this into a recommendation. Today, a growing share of that production chain can be accelerated by AI systems.

The issue is not that a chatbot can independently manage a complex corporate transformation. It cannot. The issue is that clients will find it increasingly difficult to keep paying the same price for preliminary activities that now require far less time than they once did.

Accenture: the first warning signal

On 18 June 2026, Accenture reported quarterly revenues of $18.72 billion, up 6%, but new bookings of $19.32 billion, down 2% in dollar terms year on year. The company also trimmed its expected annual revenue growth to 3–4% in local currency.¹

The market reaction was severe: the stock shed more than 17% in a single session. Reuters attributed the move in part to the impact of the conflict with Iran on Middle Eastern operations, estimated at approximately $400 million in bookings, as well as weakness in US federal spending and heightened client caution.²

The sell-off should not be read as evidence that AI is destroying Accenture. That would be an overstatement. It should be read as a signal that the market is discounting the model built on staffing, large generalist programmes, documentation and traditional consulting when demand slows and automation reduces the perceived value of standard-phase work.

Accenture also remains one of the key players called upon to implement AI, cybersecurity, cloud and automation at large corporations. The same technology can therefore compress one part of the business while fuelling another. The question for investors is which of the two forces will weigh more heavily on growth and margins.

Gartner, Globant and the value of information

The Gartner case illustrates the same dynamic from a different angle. In February 2026, the stock fell more than 20% following guidance that disappointed expectations and signs of slowing demand for advisory services. In May, however, the group reported contract value of $5.3 billion, up 1% at constant exchange rates, and earnings above consensus.³⁴

The lesson is clear: proprietary research, access to analysts and reputation continue to hold value. The mere aggregation of publicly available information, on the other hand, will become increasingly difficult to sell as an exclusive product.

Globant offers another useful data point. In the first quarter of 2026, it posted revenues of $607.1 million, down 0.7% year on year, and is seeking to shift its operating model towards subscription-based services and AI work units, known as AI Pods.⁵

This is not the end of technology services. It is the end of the comfortable equivalence between value and the number of people allocated to a project.

Where AI truly wins

The economic pressure is already visible in the most commoditised and standardisable activities. A 2026 working paper, based on corporate payments to online labour marketplaces and AI model providers, estimates that among the most exposed firms, one dollar less of spending on online labour is associated with approximately three cents of new spending on AI services.⁸

The point is not whether to use AI or not. The problem is using it without a process. A model can produce a convincing analysis, but it does not guarantee that the analysis is correct, complete, contextualised or actionable.

What the market should monitor

Sector watchlist

When the financial adviser becomes a chatbot

AI in financial advisory: opportunities and limitations

The same dynamic is reaching financial advisory. An investor can use AI to read a quarterly report, compare valuations, synthesise macro data, build scenarios and prepare a preliminary portfolio hypothesis. This represents enormous progress.

But a sound analysis is not automatically a sound investment. A public system does not necessarily know the investor's overall net worth, tax position, liquidity needs, future income, family constraints, time horizon or maximum tolerable drawdown.

ESMA has warned investors not to rely exclusively on public AI tools for decisions capable of affecting their financial well-being, as these tools may produce inaccurate, misleading or inappropriate guidance.⁶

Honesty is required here: the point is not to defend the human adviser as such. A poor human adviser is worth less than a well-deployed AI system used correctly. The true differentiator is the quality of the process: information gathering, suitability assessment, risk construction, tax planning, monitoring and accountability for the recommendation.

The limitation becomes apparent when losses arrive. An investor may follow an AI-generated suggestion, watch the stock fall 15%, and return to the chatbot asking what to do. Depending on the prompt, the system can construct a plausible response to hold, sell, or average down the position.

The problem arises earlier: if the investment was entered without a thesis, a consistent position size, invalidation conditions, and a stop-loss level, the AI's response risks becoming a post-hoc justification.

Behavioural research speaks of algorithm aversion: after observing an error, people tend to lose confidence in algorithms more quickly than they do with humans, even when the algorithm remains broadly sound.⁷

In investing, this can be fatal. You follow the machine in the initial phase, absorb a loss, and abandon the process precisely when discipline is most needed.

The old pricing model is not coming back

The most credible trajectory is not the disappearance of advisory services, but the disappearance of part of their old economic value. The production of analysis will become abundant. The ability to translate it into sound decisions, real implementation, and disciplined behaviour will remain scarce.