iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Babak Hodjat’s view is that AI can be trusted with decisions when it can act autonomously in unfamiliar situations and still produce acceptable outcomes—but that trust must be earned, monitored and shaped by explicit ethical safeguards. That perspective runs through two separate interviews: a Unite.AI interview published October 17, 2025, and The Meridian Dialogue published February 26, 2026. They are distinct conversations, not a single “return” interview.
What does Hodjat think AI should be trusted to do?
In The Meridian Dialogue, Hodjat argues that human involvement in every decision is not always necessary or desirable. People already delegate decisions to non-AI technologies, he says; whether to delegate to an AI system should depend on how trustworthy it is and on the surrounding context and culture.
He describes intelligence as the ability to make decisions that support survival in contexts not yet known. For AI, he frames trustworthiness in terms of autonomous action under unforeseen conditions that results in acceptable outcomes. This is his position, not a universal standard or a claim that current systems reliably meet it.
His proposal puts ethical responsibility across the system lifecycle: “Ethical and moral considerations should be codified unambiguously.” He says that means addressing safeguards during development and continuing to monitor systems after deployment. Delegation, in this view, is not a reason to abandon oversight; it makes clear rules and monitoring more important.
#1 Best Overall
Read The Meridian Dialogue interview, published February 26, 2026.
How does evolutionary AI differ from generative AI?
The 2025 Unite.AI interview discusses evolutionary algorithms in the context of a land-use decision-support project. Rather than treating the problem only as generating an output or predicting one outcome, the described system searches among possible policy scenarios and weighs competing objectives. The interview does not establish a quantified performance comparison with generative AI or other methods.
- Evolutionary search: explores candidate scenarios to find options that balance multiple goals.
- Generative AI: typically produces content or candidate responses from a prompt; that alone does not resolve how competing policy objectives should be prioritized.
These are broad distinctions, not a claim that the methods cannot be combined. The project description is specifically about using an evolutionary approach to explore trade-offs and support decisions.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesRank #2
What does the land-use project aim to do?
In the Unite.AI interview, Hodjat describes work with the University of Texas at Austin using Cognizant Neuro AI Decisioning technology and the Project Resilience platform. The stated aim is to help consider land-use choices that may affect carbon storage, economic disruption, food supply and habitat preservation.
The interview says the framework explores thousands of scenarios and can offer region-specific recommendations using historical land-use and carbon data. The scenarios account for contextual factors such as land type and latitude. It also describes a connection to Climate Interactive’s En-ROADS simulator to explore climate-policy incentives.
This is a description of the project and its intended decision-support role—not independent evidence of improved accuracy, implemented policy, or real-world environmental or economic results. The interview does not provide a measured outcome statistic.
Read the Unite.AI interview, published October 17, 2025.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →What can an enterprise AI-agent scoping demo show?
At Cognizant’s Investor Day on March 25, 2025, Hodjat described a prospective agent for assessing eligibility for Small Business Administration loans. The demonstration walked through identifying a possible use case, its data requirements and objectives. It illustrates early-stage scoping: how an organization might define a task for an agent before building or deploying one.
Hodjat explicitly said the example was not implemented at KeyBank. The initial analysis relied on publicly available information and guesswork, including its return-on-investment estimates. It should therefore not be treated as a customer deployment, verified business case or proof of realized returns.
See Cognizant’s March 25, 2025 Investor Day materials.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What is Hodjat’s role at Cognizant?
Cognizant’s profile identifies Hodjat as Chief AI Officer and says he leads Cognizant AI Lab. The lab’s remit includes advancing AI research, developing capabilities for Cognizant offerings and leading AI-for-good initiatives. That combination connects research to organizational applications as well as social-impact work.
His earlier career includes Dejima, Sentient Technologies and Sybase iAnywhere. Unite.AI says Dejima’s agent-oriented technology for intelligent interfaces was behind Apple’s Siri, offering historical context for his continuing interest in agents and automation. Cognizant’s profile credits him with more than 50 published papers and 39 issued U.S. patents; those biographical counts are the company’s figures.
Best Value
The titles used in sources vary by date. Cognizant’s current profile calls him Chief AI Officer; at the March 2025 Investor Day, he introduced himself as Cognizant’s “CTO AI.”
See Cognizant’s profile of Babak Hodjat.
What these conversations establish—and what they do not
Together, the interviews and Investor Day remarks show Hodjat discussing AI as both a decision-making technology and an organizational capability. His examples span multi-objective policy exploration, enterprise-agent design and the ethical conditions he believes should govern delegation.
They do not establish that the land-use project has produced measured policy outcomes, that the KeyBank example became a deployed agent, or that Hodjat’s views represent a consensus across AI developers. The distinction matters: a described capability, a scoping demonstration and a verified deployment are different kinds of evidence.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
