Maria Sanchez, senior vice-president of AI product delivery at G42’s Inception, disclosed the investment application during a panel on Wednesday, describing a system that examines proposed deals and produces arguments against proceeding with them. The fund was not identified.
The technology provides investment committees with an additional assessment of transactions supported by extensive research and documentation. Human specialists examine the AI-generated arguments for factual errors and inappropriate use of information before incorporating them into deliberations.
Sanchez said investment teams can spend months preparing material to support a transaction, only for an AI system to present reasons for rejecting it. The arrangement is intended to test assumptions rather than transfer final investment authority to software.
The discussion, titled “Powering Adaptive Economies with Agentic AI”, also featured Rima Semaan, chief solutions officer at Microsoft UAE. Kunal Devani, a partner at EY, moderated the session, held during the forum’s opening day on 14 October.
Sanchez described another application in procurement, where AI agents can assist with successive stages between purchase requests and contract awards. Tasks include checking available budgets, drafting scopes of work and suggesting criteria for assessing bids. Employees can examine, amend or reject the resulting proposals.
She advocated introducing automation first in lower-risk activities where decisions can be reversed. Organisations could then extend its responsibilities as operational experience and confidence increase, while maintaining safeguards for more consequential actions.
The distinction matters because agentic AI differs from conventional chatbots that primarily respond to questions. Agents can carry out sequences of tasks, use organisational information and interact with business systems, potentially affecting financial and contractual decisions. Such capabilities increase the importance of permissions, record-keeping and independent review.
Sanchez emphasised that companies need to reconstruct how an agent reached a decision and establish precisely when a person intervened. Traceability, she argued, is necessary if managers are to identify errors, investigate unexpected outcomes and retain responsibility for decisions.
Semaan said businesses should establish the infrastructure required for practical applications rather than indefinitely postponing implementation while awaiting more advanced models. She identified connected corporate data, suitable access controls and visibility into agents’ activities as requirements for deploying the technology across an organisation.
She described the emerging business structure as “human-led, agent-operated”, with staff directing and supervising digital agents rather than merely using AI as a tool for isolated tasks. This would require employees to develop delegation, judgement and leadership skills alongside technical familiarity.
Semaan said employees would increasingly function as managers of AI-driven work, checking whether delegated tasks had been completed appropriately and whether automated actions met organisational requirements. She also called for continued learning as systems and responsibilities evolve.
Both executives linked wider adoption to leadership involvement. Sanchez said executives who apply AI to valuable aspects of their own work can encourage its use elsewhere in their organisations. Semaan argued that companies should reconsider how work is organised and assess opportunities beyond reducing administrative effort, including possible new sources of revenue.
The discussion comes as Microsoft and Inception pursue a collaboration announced in July to connect their agentic AI offerings for government and enterprise customers. The companies said the initiative aims to support coordinated deployment and governance while accommodating requirements for domestic data processing.
A separate partnership between Inception and Abu Dhabi’s Department of Government Enablement, announced in February, includes AI applications for procurement and human resources. Its procurement plans cover supplier registration, purchasing assistance and preparation of requests for proposals, illustrating the types of administrative functions being targeted.
The examples presented at the forum did not establish how much time or money the organisations had saved, and no performance figures were disclosed for the investment-review system. The executives focused instead on operational controls and the gradual allocation of responsibilities. Their accounts also distinguished AI-generated recommendations from decisions made by personnel, an important separation where financial commitments and procurement awards are involved.
The ninth Sharjah Investment Forum, organised by the Sharjah FDI Office in collaboration with the UAE Ministry of Investment, runs on 14 and 15 October at Al Jawaher Reception and Convention Centre. Its programme covers economic adaptability, investment flows, artificial intelligence and supply-chain resilience.
Follow Arabian Post
Select Arabian Post as your preferred source on Google and MSN News for trusted business news and Arab politics and updates.
