ibl.ai Agentic AI Blog

Insights on building and deploying agentic AI systems. Our blog covers AI agent architectures, LLM infrastructure, MCP servers, enterprise deployment strategies, and real-world implementation guides. Whether you are a developer building AI agents, a CTO evaluating agentic platforms, or a technical leader driving AI adoption, you will find practical guidance here.

Topics We Cover

Featured Research and Reports

We analyze key research from leading institutions and labs including Google DeepMind, Anthropic, OpenAI, Meta AI, McKinsey, and the World Economic Forum. Our content includes detailed analysis of reports on AI agents, foundation models, and enterprise AI strategy.

For Technical Leaders

CTOs, engineering leads, and AI architects turn to our blog for guidance on agent orchestration, model evaluation, infrastructure planning, and building production-ready AI systems. We provide frameworks for responsible AI deployment that balance capability with safety and reliability.

Back to Blog

How Should a State Agency Procure AI Agents It Can Own and Self-Host?

ibl.aiNovember 18, 2026
Premium

A state agency buying AI agents it can own runs six steps: define the use case, assess risk against NIST AI RMF or the state's own tool, pick a vehicle such as NASPO ValuePoint, clear the state's cloud security program, write ownership and data terms, and keep monitoring.

The Short Answer

A state agency procures AI agents it can own by treating ownership as a contract term, not a feature: assess risk, buy through an approved vehicle, clear the state's cloud security program, then require source code, data return and model switching in writing. On ibl.ai you own all the code and the data, and the agency self-hosts it.

The six steps below follow published state and national guidance: NIST's AI Risk Management Framework, California's GenAI procurement process, NASPO ValuePoint and Texas DIR's TX-RAMP program.

The rules differ by state, so treat this as the shape of the process and confirm each step with your state's procurement office.

What are the steps for a state agency to procure AI agents?

A state agency procuring AI agents typically works through six steps, and California's published process alone has five stages from training to ongoing monitoring.

StepWhat the agency doesPublished example
1. Define the use caseName the workflow, the systems the agent reads, and the data classificationCalifornia starts with GenAI training before the procurement package
2. Assess riskRate the use against a risk frameworkNIST AI RMF 1.0; California's SIMM 5305-F risk assessment
3. Pick a vehicleUse a cooperative or state contract, or run a solicitationNASPO ValuePoint Cloud and Software Solutions, 2026 to 2036
4. Clear security authorizationMeet the state's cloud security programTexas TX-RAMP Level 1 or Level 2
5. Write ownership and data termsSource code, data return, model switching, logsCalifornia requires GenAI disclosure and reporting language
6. Monitor after go-liveAssign a contract manager and review the agent's behaviorCalifornia requires a contract manager for every GenAI contract

The order matters. Ownership terms written after a vehicle is chosen are limited to what that vehicle's master agreement allows, so step 5 has to inform steps 3 and 4.

How should a state agency assess AI agent risk before buying?

A state agency should assess AI agent risk against a recognized framework before writing the solicitation. The common baseline is NIST's AI Risk Management Framework 1.0, released January 26, 2023, which NIST says "is intended for voluntary use."

NIST added a Generative AI Profile, NIST AI 600-1, on July 26, 2024, and says AI RMF 1.0 is being revised under the White House AI Action Plan.

The framework's four functions, Govern, Map, Measure and Manage, give an agency a vocabulary for the risk file even where state law does not require it.

Some states require their own tool. California entities assess GenAI risk with SIMM 5305-F, and consultation with the California Department of Technology is required for moderate- and high-risk procurements and optional for low-risk ones.

For AI agents the risk turns on actions, not just answers. An agent that reads case files or writes to a system of record needs its permissions, its data paths and its logs in the risk file, alongside the model.

Which contract vehicles can a state agency use to buy AI agents?

A state agency can buy AI agents through a cooperative contract, an existing statewide contract, or its own solicitation.

One option is newly awarded: NASPO ValuePoint's Cloud and Software Solutions portfolio, led by Utah, was awarded September 16, 2026 and expires September 15, 2036.

The portfolio features 58 awarded suppliers and covers SaaS, IaaS, PaaS, on-premise software and value-added services. That last pair matters for an agency that wants to self-host: the vehicle is not limited to vendor-hosted SaaS.

A cooperative contract only works where the state has signed a participating addendum. NASPO says each addendum "outlines which entities within the state" can purchase; without one, the agency goes to the state's chief procurement official.

Federal vehicles are a separate track. For GSA MAS, OASIS+ and similar routes, see our guide to government AI contract vehicles.

What security authorization does a state agency's AI agent need?

A state agency's AI agent needs whatever cloud security authorization the state requires, and Texas shows how binding that can be.

Under Government Code Section 2054.0593, Texas state agencies may only enter or renew contracts for cloud computing services that comply with TX-RAMP, beginning January 1, 2022.

TX-RAMP has two levels: Level 1 for low impact information resources and Level 2 for moderate or high impact ones, per version 3.0 of DIR's program manual. Each cloud service needs its own certification.

One detail corrects a common assumption.

The manual exempts bespoke applications an agency commissions, but says the exemption "does not include commercial off-the-shelf software platforms that are simply configured specifically for agency purposes." A configured AI platform is not exempt by default.

Self-hosting changes where the boundary sits rather than removing it. If the agency runs the software in an environment it already operates, the scoping question is about that environment; confirm the determination with DIR or your state's equivalent and document it.

What ownership and data terms should a state AI agent contract include?

A state AI agent contract should put ownership in writing, because the vehicle and the security review do not supply it. California already requires GenAI disclosure language in every written solicitation and GenAI reporting language in every written contract.

Beyond disclosure, five terms decide whether the agency owns its agents or rents them.

Contract termWhat it should say
Source codeThe agency receives the platform's source code under a perpetual license
Data returnAll data, prompts, transcripts and configurations exportable in open formats at no extra charge
Model switchingThe agency can change LLM provider without a contract amendment
LogsAudit logs and transcripts are held in the agency's environment
PricingUsage-based against a cap the agency sets, not per seat

The federal government has moved in the same direction: OMB's M-25-22 asks AI contracts for data portability and protection against lock-in. We covered that memo and a GAO finding on exit costs in government AI procurement moving to architecture ownership.

What should a state agency monitor after its AI agents go live?

A state agency should monitor its AI agents continuously after go-live, because an agent's behavior changes when its model, prompts or connected systems change.

California requires every GenAI contract to have a contract manager who continuously oversees and evaluates the technology for risk.

California also covers new capabilities on an existing contract: GenAI added through a contract amendment still requires a risk assessment. Each state entity must report all GenAI purchases into FI$Cal.

Monitoring works only on what the agency can see. On ibl.ai that means transcripts of every agent conversation, an audit log of configuration changes, roles and RBAC scoping each agent, and cost analytics per agent.

Where does ibl.ai fit in a state agency's AI agent procurement?

ibl.ai fits step 5 as the owned-stack option: the agency receives the code and runs the platform in its own environment.

On ibl.ai you own all the code and the data, run it model-agnostic across any LLM, with no per-seat pricing, and deploy anywhere from a state data center to a fully air-gapped network.

ibl.ai is SOC 2 Type II certified. It does not replace the state's own steps: the agency still runs its risk assessment, uses its vehicle, and completes its state's security determination.

ibl.ai is family-owned and operated from New York, NY, and its engineers build alongside an agency's own team, so the agents the state commissions stay the state's.

Sources: NIST AI Risk Management Framework (AI RMF 1.0, January 26, 2023; NIST AI 600-1, July 26, 2024); California GenAI procurement process (California Department of Technology); NASPO ValuePoint Cloud and Software Solutions; Texas DIR TX-RAMP and its Program Manual v3.0. DIR has since published a v3.1 manual; check the current version before relying on section numbers.

Why does owning the AI stack matter?

ibl.ai is the agentic AI platform where you own all the code and the data. You self-host the entire stack inside your own perimeter, run it model-agnostic across any LLM and switch anytime, and pay by usage with no per-seat pricing β€” so you can deploy anywhere: your cloud, on-premise, GovCloud, or fully air-gapped.

  • You own all the code and the data

    Full source code under a perpetual license, running on your infrastructure. Not API access to someone else's platform β€” the stack itself is yours.

  • Model-agnostic

    Run any LLM β€” Claude, GPT, Gemini, Llama, Command, or your own fine-tune β€” and switch providers without rewriting the platform.

  • No per-seat pricing

    Usage-based billing against a budget cap you set. Cost tracks what your organization actually uses, not how many people you employ.

  • Deploy anywhere

    Your cloud, your VPC, on-premise, GovCloud, or a fully air-gapped network with no outbound connectivity.

1.6M+ users across 400+ organizations run the platform this way, including NVIDIA, MIT, and Syracuse University.

ibl.ai is family-owned and operated from New York, NY β€” a U.S.-headquartered, domestically-owned long-term partner, not a vendor that sells licenses and moves on.

Related Articles

Craft Anything: When an Agent Can Rebuild the App, What Is Still Worth Owning?

Craft Anything is an open-source, pure-Rust engine from the ArtCraft team that studies a website, app or game and has a coding agent build and check a new version. It passed 25,000 GitHub stars ten days after its repository opened. When the interface can be rebuilt, the asset worth owning is your data and the relationships in it.

Miguel AmigotNovember 10, 2026

Government AI Procurement Is Moving From Vendor Selection to Architecture Ownership

OMB's M-25-22 asks AI contracts for data portability, lock-in protection and rights to code produced under the contract, and GAO found a contractor charging an agency to get its own data back. What architecture ownership adds, layer by layer, and the contract clauses that make it real.

ibl.ai EngineeringNovember 9, 2026

Germany Now Owns the Source Code of Its Combat-Drone Mission Software. What the Helsing MissionOSBw Handover Means for AI Procurement

Helsing handed Germany's procurement agency the full source code, documentation and SDK for MissionOSBw, with ownership staying with the German state. What was transferred, what is still unclear, and the procurement lesson for any agency buying AI.

ibl.aiNovember 7, 2026

Eleven Coding Agents, One Architecture: What a Source-Code Study Means for Enterprise Teams

A Wavestone AI Lab study of eleven coding agents' source code found they converge on seven subsystems, use no vector search for code, and differ most in safety and permissions, which is where enterprise teams should spend their evaluation.

Jaione AmigotNovember 6, 2026

See the ibl.ai AI Operating System in Action

Discover how leading universities and organizations are transforming education with the ibl.ai AI Operating System. Explore real-world implementations from Harvard, MIT, Stanford, and users from 400+ institutions worldwide.

View Case Studies
Work with our team

Pilots, deployment, and full ownership

Every engagement is scoped to the work: you integrate ibl.ai with your own data, deploy it on your own infrastructure, and the engineering hours scale with the scope β€” so the price tracks the scope, not your headcount. Most customers keep us on afterwards for platform updates, new agents and integrations, and ongoing services.

Start here

Pilot

from $15K

fixed scope Β· fixed timeline

A time-boxed proof of value on your real data β€” not a slide deck.

Best for: Teams that want to see ibl.ai working before committing.

  • Deployed on your infrastructure or our cloud
  • 1–2 production agents wired to a slice of your data
  • One integration (LMS / SIS / SSO / data source)
  • Weekly working sessions with our engineers
  • Pilot fee credits toward a full engagement
Scope a pilot
Most common

Integration & Deployment

$25K – $80K

one-time fee Β· scoped to the work

Full deployment integrated with your data and systems. Engineering hours scale with scope.

Best for: Organizations rolling ibl.ai out across a department, campus, or business unit.

  • Platform deployed in your VPC, on-prem, or air-gapped
  • Integrated with your data + identity (SSO / SAML)
  • Multiple custom agents built to your workflows
  • Engineering hours proportional to scope
  • You own the data Β· run any LLM you choose
Plan a deployment
Full ownership

Codebase Transfer + Custom AI Engineering

Custom quote

perpetual license Β· you own the stack

We transfer the full source code. You own and self-host the entire platform β€” outright.

Best for: Organizations and enterprises that benefit from perpetual ownership and sovereignty.

  • Complete source-code transfer + perpetual license
  • Dedicated AI engineering team on your roadmap
  • Custom agents, models, and integrations to spec
  • Air-gapped capable Β· zero vendor lock-in
  • Family-owned, New York–based long-term partner
Talk about ownership
You own the code and data Run any LLM β€” Claude, GPT, Gemini, Llama Family-owned & operated from New York, NY