The adoption of artificial intelligence is accelerating the transformation of technological hiring models by companies. The proliferation of AI initiatives, as well as the growing deployment of AI agents and their integration into traditional IT solutions, requires organizations to review how they select, contract and manage these services, according to the report ‘IA Sourcing: the transformation of the IT services contracting model’, prepared by Deloitte.
Currently, 75% of companies plan to deploy autonomous AI agents in the next two years.
Artificial intelligence has ceased to be a tool aimed exclusively at efficiency and has become an engine of business transformation. This evolution is reflected in the fact that half of employees already have access to AI tools and that 34% of organizations have placed it at the center of their transformation strategy. Additionally, many companies are already exploring the use of AI agents to automate complex processes and improve decision making.
As these solutions and AI agents gain relevance within companies, exposure to operational, cybersecurity, financial, reputational and strategic risks also increases. In parallel, AI initiatives are leaving behind the pilot and proof-of-concept phase to focus on programs with a direct impact on operations and business strategy, driven in many cases by specialized AI agents.
Javier Martín Barroso, partner of Tech Strategy at Deloitte, points out that “the mass adoption of artificial intelligence is redefining the rules of technological contracting. The organizations that obtain a competitive advantage will be those capable of integrating AI and AI agents within a solid framework of governance, risk and value creation that allows them to scale their impact in a sustainable and responsible way.”
The cost of AI deployment increases
The Deloitte report reveals that the reduction in unit costs of artificial intelligence is not translating into lower investment by organizations. In fact, the total cost associated with AI solutions and the deployment of AI agents has increased between 50% and 75% in the last year, despite the fact that the cost of tokens has reduced up to 280 times in just two years.
This increase responds to various factors, including the proliferation of initiatives driven in a decentralized manner by business areas, the implementation of AI agents in multiple departments, the absence of sufficiently mature governance models, the need to incorporate new internal capabilities, the adaptation of operating models and the limitations of certain legacy technological architectures.
This complex ecosystem has direct consequences on results, since 30% of AI initiatives do not achieve success, mainly due to the high associated costs and the lack of specific models for the operation and management of systems based on AI agents.
Redefinition of technological contracting models
Traditional sourcing models are no longer valid to manage technologies with a higher level of autonomy, new risk profiles and a growing dependence on data. This is especially relevant given the emergence of AI agents, capable of executing tasks, making decisions and coordinating processes autonomously. Therefore, AI sourcing should be treated as a complex process that integrates technological, legal, ethical, operational and organizational dimensions.
For Javier Martín Barroso, partner of Tech Strategy at Deloitte, “artificial intelligence providers must contribute much more than technology. Their true value lies in their ability to industrialize use cases, deploy AI agents, accompany the transformation of operating models and guarantee solid and transparent governance of these solutions. In this context, contracting models must evolve to incorporate in a structured way the management of the new risks associated with AI and AI agents, enabling scalable, secure adoption aligned with business objectives.”
To respond to this scenario, Deloitte proposes a comprehensive model for contracting artificial intelligence services, which is structured in eight phases and covers the entire life cycle of AI initiatives and AI agents:
- Definition of the use case and assumed risk: effective and sustainable use of AI.
- Internal maturity and readiness evaluation: use case sheet with expected value and approved risk level.
- Strategy and governance: buy “as-a-service”, co-create or develop internally.
- Market analysis and supplier shortlist: specific governance model and approved solution strategy.
- Technical, ethical and legal evaluation: shortlist of suppliers.
- Contractual design and economic model: multidimensional scoring, not just technical.
- Pilot, validation and scaling: AI risk translated into clear contractual clauses.
- Operation and continuous improvement: deployment and scaling based on value and risk.
Likewise, the report highlights the need to carry out five priority actions by organizations:
- Define the bases of a comprehensive AI sourcing framework, which analyzes AI provision services and all those IT services that incorporate AI in an embedded way, including AI agents.
- Define a comprehensive and transversal AI governance model, with participation from IT, business, legal, compliance, risks, HR and security, before scaling AI pilots or agents to production.
Training for all business areas
Select suppliers aligned with the transformation strategy, not just with technical capabilities, demanding demonstrable experience in real use cases, scalable and supported by AI agents when necessary.
Deloitte proposes a comprehensive model for contracting artificial intelligence services
Evaluate and choose pricing models that balance predictability and scalability, especially in environments with intensive use of AI agents, progressively evolving towards results-oriented schemes as adoption matures.
In a context marked by the growth of artificial intelligence and the deployment of increasingly autonomous AI agents, it will be key to have a structured contracting model to take advantage of the potential of these technologies while maintaining control over the associated risks.
