By Global Consultants Review Team ,
The consulting industry is entering a harder buying cycle. Generative AI and AI agents can now perform more research, coding, analysis, documentation and workflow tasks inside client organizations.
That is reducing demand for some repeatable work while increasing demand for implementation, integration, governance and organizational change.
AI consulting is not disappearing. Clients are changing what they buy, how they measure value and what they are willing to pay.
On September 1, the Financial Times reported that companies including Bayer and Bristol Myers Squibb are using AI to bring more technology work in-house. The report said buyers are putting pressure on consulting fees as AI improves internal capacity for areas such as system integration and cybersecurity. (Financial Times, 2026)
The pressure sits beside continued demand for large transformation programs. Accenture reported third-quarter fiscal 2026 revenue of $18.7 billion and new bookings of $19.3 billion. Chief Executive Julie Sweet said the company was seeing more large-scale AI transformation programs.
These signals are not contradictory. Companies can spend less on routine consulting while spending more on difficult AI programs that require integration, security and operating-model changes.
Buying criteria are changing as well. BCG reported in February that more than 70 percent of enterprise decision-makers in its technology-services research preferred output- or outcome-linked commercial models for customer experience and business process outsourcing services. Yet nearly 60 percent said they had not seen measurable total-cost improvement in deals containing agentic AI.
Consulting firms are responding by moving closer to technology platforms and implementation. BCG, for example, expanded its Google Cloud partnership in April around Gemini Enterprise transformation and agent deployment tied to measurable business results.
Internal teams already know their company's data, applications and operating limits. That knowledge becomes important when AI touches customer information, intellectual property, cybersecurity or regulated processes.
AI development has also become more accessible. Foundation models, cloud AI services and coding agents allow internal technology teams to test ideas without building every component themselves.
McKinsey's August 2026 State of AI survey found that 32 percent of respondents said their organizations had decided against buying at least one software product or feature because they could build it internally with agentic coding tools. Among large organizations, 40 percent reported scaling AI agents.
Internal development can also improve speed. Business and technology teams can test a workflow, measure its value and change it without managing a large external project.But insourcing has limits. Companies still need help when data is fragmented, systems are difficult to integrate or AI affects many business functions.
McKinsey reported in April that data limitations remain a major barrier to scaling AI agents. Its August research on organizational adoption also argued that AI transformation depends heavily on process redesign, skills and changes in how people work.
Repeatable tasks with clear inputs and standard outputs face the strongest pressure.
AI can help produce first drafts of market scans, competitor reviews, presentations, process maps, reports and basic technology assessments. Coding assistants can handle portions of software development and testing.
Research-heavy work is also changing. A consultant may still need to validate evidence and draw conclusions, but gathering and summarizing information can take far less time.
Complex problems remain different. Decisions involving regulation, organizational politics, investment trade-offs or unclear evidence still require human judgment.
This suggests a divided consulting future. Commodity analysis faces stronger price pressure. High-stakes judgment, specialized knowledge and execution remain harder to replace.
Traditional consulting economics often connect revenue to hours, team size and employee seniority. AI challenges that relationship when a smaller team can produce the same output more quickly.
This pressure is already visible in India's IT-services sector. Reuters reported in August that TCS, Infosys, Wipro, HCLTech and Cognizant are adjusting business models as customers demand higher productivity and lower prices.
TCS Chief Executive K Krithivasan told Reuters that about 80 percent of contracts in the company's finance, human resources and other business-services segment are now based on outcome performance measures.
Reuters also reported that some customers are taking work in-house and using shorter contracts. Infosys had told analysts that it walked away from contracts that were no longer economically viable.
The result is growing interest in fixed fees, milestone payments, subscriptions, retainers and shared-savings arrangements alongside traditional billing.
AI outcome-based pricing changes the conversation. Instead of paying mainly for how many people work on a project, the buyer can connect part of the fee to an agreed result.
For example, a technology provider could receive payments when an AI application reaches agreed deployment milestones. A managed-service contract could connect fees to service performance. A savings contract could share verified efficiency gains.
The difficult issue is measurement. Both parties must agree on the starting point, the expected result and which factors the consultant can actually control.
Clients increasingly want working systems rather than generic AI strategy reports.
They expect faster prototypes, secure deployment and integration with existing applications. They also expect training, governance and knowledge transfer so internal teams can continue operating the system.
The demand for evidence is becoming stronger. PwC's 2026 CEO research found that only one in eight CEOs reported both higher revenue and lower costs from AI. More than half reported neither higher revenue nor lower costs.
Costs are receiving more attention as AI moves into production. KPMG reported in June that only 26 percent of organizations in its US survey had real-time visibility into AI operating costs. EY reported in July that 82 percent of surveyed senior leaders whose organizations were investing in AI were concerned about token usage and related costs. Tokens are the units used to measure how AI models process and generate information.
For procurement leaders, the central question is changing from "Can you build AI?" to "What measurable result will this project produce, at what cost?"
Consultants retain an advantage when problems cross departments, technologies and executive interests.
Regulation, cybersecurity, operating-model design and large transformations require more than information generation. They require choices, accountability and agreement among people with competing priorities.
Independent judgment also matters. Senior executives sometimes hire advisers because they need an external view of difficult decisions rather than another software tool.
Change is another major area of demand. McKinsey's 2026 research argues that moving agentic AI beyond pilots requires substantial work on processes, skills and adoption. Technology alone does not change how an organization operates.
Consultants must also use AI themselves. A firm selling AI transformation cannot defend traditional fees if it still performs research, coding and documentation at pre-AI speed.
India's IT-services firms face direct pressure because much of the sector grew through large delivery teams and long outsourcing contracts.
Reuters' August reporting showed that clients are seeking measurable outcomes and stronger productivity commitments. It also found that employee scale is becoming less decisive as AI automates more work.
The same report said TCS was increasing the number of engineers embedded with clients to accelerate AI adoption. That points toward a different talent model: fewer advantages from raw headcount and greater value from people who can combine AI, engineering and business knowledge.
Demand can shift toward AI engineering, integration, cybersecurity and managed AI services. Indian providers also have an opportunity to replace labor-based contracts with agreements linked more closely to business results.
Global capability centers create another challenge. As multinational companies expand their own technology and AI teams in India, some work that might once have gone to an outsourcing provider can remain inside the company.
The key question is whether new AI-related revenue grows faster than routine work is automated or insourced.
Over the next two to five years, smaller AI-enabled consulting teams are likely to handle more work. This could reduce billable hours even when the business value of a project remains high.
More clients are also likely to maintain permanent internal AI teams. Outside consultants may become more concentrated around difficult transformations, specialist expertise and periods when clients need additional capacity.
Fixed-fee, milestone and outcome-linked contracts should become more common where results can be measured. Shared-savings agreements may also grow, although disputes over baselines and attribution could limit their use.
Specialist consulting firms may gain opportunities in AI security, governance, data engineering and industry-focused implementation. Larger firms are likely to keep responding through acquisitions and partnerships with AI and cloud providers.
Procurement teams may also demand greater transparency about how consultants use AI. Buyers will want to know whether automation has reduced required staffing, how client data is protected and whether efficiency gains are reflected in AI consulting fees.
We use cookies to ensure you get the best experience on our website. Read more...
