HIPAA-compliant AI consulting for healthcare operations: compare your options in 2027
A consulting firm that touches PHI during discovery, data mapping or builds is a HIPAA business associate and must sign a BAA before work starts; tool-vendor BAAs do not cover it. The BAA chain runs provider, consulting firm, model provider, cloud host, each hop on a BAA-eligible tier.
HIPAA-compliant AI consulting requires the firm itself to sign a BAA, map every PHI data flow, and complete a risk analysis before any build. Every vendor in the chain, from model provider to cloud host, needs its own agreement on a BAA-eligible service tier. Compliance is tested where the system is built, which is where OCR is looking.
By March 2026, HHS's Office for Civil Rights had announced 12 enforcement actions under its Risk Analysis Initiative, the latest a settlement with a Maryland software company over a breach affecting 15 million people. Every action in the initiative turns on the same requirement, an accurate and thorough risk analysis of where electronic PHI lives and moves. For a multi-site medical group or community hospital starting an AI project, that analysis has to cover the implementation layer, meaning the systems, data flows, and vendors the project actually touches.
Most guidance on HIPAA-compliant AI covers how to buy software, with BAA checklists for tool vendors. This article covers the question those guides skip, which is what a compliant AI consulting requires when a firm's people, not just its tools, work inside your PHI environment.
---
Key takeaways
- This list is for healthcare providers and business associates with US$10M–$1B in annual revenue, such as multi-site medical groups, community hospitals, health plans, and digital health companies, that are starting an AI project that touches PHI.
- Each firm was screened on four criteria. No firm paid to appear on this list.
- Healthcare focus. The firm's compliance, governance, or delivery work serves healthcare organizations.
- BAA position. The table states whether each firm signs a BAA, or that this is not stated.
- Build and operate roles. The table states who builds the AI and who operates it.
- Provider-size fit. Each firm is placed by the size of provider it serves.
- A consulting firm that touches PHI during discovery, data mapping, or builds is a business associate under HIPAA and must sign a BAA before the work starts, and tool-vendor BAAs do not cover it. No AI model or tool is HIPAA compliant on its own, because compliance comes from a BAA with every party that handles PHI, a risk analysis that includes the AI system, and controls on how PHI moves through the build.
- OCR's Risk Analysis Initiative reached 12 enforcement actions by March 2026, all centered on risk analysis, and healthcare breaches average $6.64 million each, the costliest of any industry for the 13th straight year per IBM.
- Where PHI flows through a workflow, Boldr AI puts the BAA and the data-flow map ahead of the build. Sensitive information stays with people.
HIPAA-Compliant AI Consulting, Defined
HIPAA-compliant AI consulting is work in which the consulting firm operates as a business associate, with a BAA signed before any PHI is accessed, compliance controls built into the delivery process itself, and the AI systems delivered in a configuration the client can defend in an audit. The defining feature is that the firm's own conduct is regulated, covering how its people handle your data in workshops, notebooks, test environments, and logs. "Our tools have BAAs" describes a software stack, and it says nothing about the humans doing discovery in your scheduling system.
This guide covers operations AI, meaning scheduling, intake, revenue cycle, documents, and back-office workflows. Clinical AI, the world of diagnostic and treatment algorithms, carries a different risk class, FDA exposure, and a different buyer. Operations work is where multi-site medical groups, independent groups, and community hospitals get faster payback at lower regulatory risk, and it is the lane this guide stays in.
The same requirements reach beyond providers. Health plans, third-party administrators, and digital health companies are covered entities or business associates in their own right, so everything below, from the BAA chain to the build controls, applies to their AI consulting projects as well.
---
What Makes AI HIPAA Compliant
No AI model or tool is HIPAA compliant on its own. HHS's Office for Civil Rights says it does not endorse, certify, or recommend specific technology or products, so a vendor's "HIPAA compliant" label is a claim about how the vendor will handle your data and carries no regulator's certification. What can be compliant is how a covered entity or business associate uses a tool, and that comes down to four conditions.
- A business associate agreement with every party that handles PHI for you. HHS's cloud computing guidance treats a provider that creates, receives, maintains, or transmits electronic PHI on your behalf as a business associate, even when it stores only encrypted data and holds no key.
- A risk analysis that includes the AI system. The Security Rule requires an accurate and thorough analysis of the risks to electronic PHI, and it is the requirement OCR's Risk Analysis Initiative enforces. A new model, integration, or vendor changes the answer.
- Access limited to the minimum PHI the task needs. The Privacy Rule requires covered entities to make reasonable efforts to limit uses, disclosures, and requests of PHI to the minimum necessary for the purpose.
- De-identification only by a method HHS recognizes. Under HHS's de-identification guidance, data counts as de-identified through Safe Harbor, which removes 18 listed identifiers, or through Expert Determination, in which a qualified expert documents that the risk of identification is very small.
---
The BAA Chain: Who Signs What in a Custom AI Build
A custom AI build creates a chain of business associates, and the chain is only as compliant as its weakest agreement. The provider group sits at the top as the covered entity. The consulting firm signs a BAA with the provider the moment its work involves PHI, which in practice means before discovery, since data mapping and workflow analysis expose PHI almost immediately.
The chain continues below the consultant. The model provider, whether OpenAI, Anthropic, or Google, has to sign a BAA that covers the specific service and plan the build uses, and each provider's own documentation says which plans qualify. The cloud host running the application, typically AWS, Azure, or Google Cloud, signs its own BAA covering the infrastructure.
Account ownership decides who signs each of those downstream agreements. Where the consultant contracts these services on the client's behalf, subcontractor BAAs flow the obligations down, and where the client holds the accounts, the client's own agreements govern.
| Hop | Agreement | Who typically signs | |----------------------------------|-----------------------------|---------------------------| | Provider → consulting firm | BAA, before discovery | Both, at project start | | Consulting firm → model provider | BAA covering the plan in use | Whoever holds the account | | Consulting firm → cloud host | BAA covering infrastructure | Whoever holds the account | | Any subcontractors | Downstream BAA | Consulting firm |
Multiple BAAs on one project is the normal, correct shape, and a candidate firm should be able to draw this chain for your specific build unprompted. Each business associate holds direct regulatory exposure for its own conduct, which is why a serious firm treats its own compliance as self-interest.
---
What Healthcare Providers and Business Associates Should Know Going Into 2027
Breach patterns, enforcement counts, and the Security Rule timeline set the compliance baseline for an AI project in 2027. The figures below come from HIPAA Journal's analysis of OCR data, a law-firm summary of OCR settlements, and Fierce Healthcare's reporting on HHS's rulemaking timeline.
- Business associates sit in a growing share of breaches. Business associates were involved in 43% of large healthcare breaches reported in the first half of 2026, against an average of 20% between 2009 and 2017, according to the HIPAA Journal's analysis of OCR data, as reported by Insurance Business. HIPAA Journal's own breach statistics page counts a record 772 large breaches in 2025, with hacking and other IT incidents accounting for more than 80% of large breaches, per HIPAA Journal.
- OCR enforcement keeps centering on risk analysis. The Spencer Gifts benefit-plan settlement of June 2026, which carried a $450,000 payment and a two-year corrective action plan, was OCR's 14th enforcement action under its Risk Analysis Initiative, according to Davis Wright Tremaine's October 2026 summary. The firm notes that failing to conduct a risk analysis is among the most frequently cited violations, which is why an AI project's risk analysis comes first.
- The Security Rule overhaul is now a 2027 event. HHS moved its target for the final amended rule from May 2026 to July 2027, according to Fierce Healthcare, which notes that the rule is 23 years old and the overhaul would be its first major update in more than a decade. Until it lands, the current rule is the one a 2027 AI project is audited against.
HIPAA-Compliant AI Vendors and Consulting Firms Compared
The market for this problem sorts into three categories, and most buyers end up needing one from each. The comparison that matters is who signs a BAA, who builds, who operates, and what provider size each serves.
Compliant infrastructure and model access. Microsoft's Azure OpenAI Service and OpenAI's enterprise offerings, Google Cloud's healthcare stack, and Anthropic's Claude on plans covered by a BAA are the rails a compliant build can run on, with BAA terms that vary by service and plan. They provide the substrate, and everything this article describes still has to be built on top of it.
1. Boldr AI
Boldr AI is an AI consulting and execution partner that fixes the process first, then automates what pays back, in systems the client owns. It serves healthcare providers and business associates with US$10M–$1B in annual revenue, where operations work is labor-heavy and PHI touches the workflow. Delivery is US-based, with senior nearshore delivery teams in LATAM working the client's business hours.
The first step is the Value Discovery Sprint, a diagnosis and design of the workflow that takes 2–4 weeks, opens with the BAA, and carries the risk analysis scope and the PHI data-flow map. It starts at US$7,500 for US work as a one-time fee, deployment is quoted on the scope the sprint defines, and the Composed POD operates the system and keeps the documentation current from US$5,000/month.
2. Medcurity
Medcurity is a HIPAA compliance software platform with advisory support, strong on risk analysis, security risk assessments, and OCR-readiness, and it packages risk analysis for smaller organizations. It will tell you where your gaps are with real authority.
Building and operating AI systems sits outside its business, so it pairs naturally with a delivery firm. It suits small practices that need a packaged risk analysis in hand before an AI project starts.
3. Clearwater
Clearwater is a healthcare compliance consultancy, strong on risk analysis, security risk assessments, and OCR-readiness, and it serves hospitals and health systems, physician groups, health plans, digital health companies, and critical access hospitals. Like Medcurity, it will tell you where your gaps are with real authority.
Building and operating AI systems also sits outside its business, so it pairs naturally with a delivery firm. Hospitals, health systems, and physician groups that want an authoritative risk analysis and OCR-readiness review ahead of an AI build are its closest match.
4. BD Emerson
BD Emerson's healthcare AI governance service is built for health systems, physician groups, payers, health tech companies, and revenue cycle vendors. It covers an inventory of the AI in use, including shadow AI, a BAA and vendor review that looks at training rights, retention, subcontractors, data location, breach-notice timing, and excluded features, and a HIPAA risk analysis alongside NIST AI RMF work. It also sets up a governance committee and charter, maps regulations, and prepares organizations for ISO/IEC 42001, with an accredited body performing the certification audit.
The work starts with a scoped inventory and BAA review, and the page lists no pricing. Because the service centers on governance, the build would come from a delivery firm, and it suits organizations that need AI governance and vendor review in place first.
5. BerryDunn
BerryDunn's HIPAA compliance consulting covers gap analysis, security risk analysis, training, and policies, with mock OCR audits, combined SOC 2 and HIPAA audits, and outsourced privacy officer support and OCR complaint and breach response. It serves covered entities and business associates across Medicaid agencies, higher education, healthcare, nonprofits, community health centers, and rural health clinics.
The page describes compliance assessment and support, not AI delivery, so it pairs with a delivery firm. It suits community health centers, rural health clinics, and other providers that want audit readiness and outsourced privacy support around an AI project.
6. Petronella
Petronella is a cybersecurity-first consultancy offering healthcare AI services with published assessment pricing, oriented toward medical practices and compliance services such as risk assessment, BAA review, and audit logging, with no standalone process-redesign service. It suits medical practices that want published assessment pricing and compliance services around an AI project.
7. CitiusTech
CitiusTech is an enterprise health IT firm with deep engineering benches and health-system clients, sized for large enterprises. It suits health systems that need engineering depth at program scale.
| Firm / platform | Signs a BAA | Builds AI | Operates it | Provider-size fit | Best for | |---|---|---|---|---|---| | Azure OpenAI / OpenAI, Google Cloud, Anthropic | Platform layer | No | No | All (infrastructure) | Builds that need BAA-covered infrastructure and model access | | Boldr AI | Yes, before discovery | Yes | Yes, Composed POD | US mid-market companies with US$10M–$1B in annual revenue | Provider organizations that want the BAA, risk analysis scope, and build under one partner | | Medcurity | Not stated | No | No | Smaller organizations | Small practices that need a packaged risk analysis | | Clearwater | Not stated | No | No | Hospitals, health systems, physician groups, health plans, and digital health (no size stated) | Hospitals, health systems, and physician groups that need an authoritative risk analysis and OCR-readiness review | | BD Emerson | Not stated | Not described | Not described | Health systems, physician groups, payers, health tech, and revenue cycle vendors | Organizations that need AI governance and vendor review in place | | BerryDunn | Not stated | Not described | Not described | Covered entities and business associates | Providers that want audit readiness and outsourced privacy support | | Petronella | Yes | Yes | Limited | Medical practices | Medical practices that want published assessment pricing | | CitiusTech | Not stated | Yes | Enterprise programs | Health systems | Health systems that want engineering depth at program scale |
---
PHI in the Build: Training Data, Logs, and De-Identification
Delivery work creates PHI exposure in places procurement checklists never look. Evaluation datasets get assembled from real appointment records, and prompt and completion logs capture patient names inside debugging traces. A developer reproducing a bug pastes a real referral into a test environment, and the environment retains it.
A compliant project names each of these paths in the risk analysis and closes them with technical controls. Policy language alone closes none of them.
Boldr AI's approach follows one house rule. As a rule, sensitive information stays with people and workflows do the mechanical part. Where PHI does flow through a workflow, a BAA covers it before the work touches it, and the data-flow map records which fields the automation may read, write, and log.
Where a build uses de-identified data, the data has to meet one of the two HHS methods described earlier, and the build environment keeps its access controls even then. A firm that can walk through these controls unprompted is operating at engineering-level compliance, which is the level audits test.
---
What OCR Enforcement Means for Your AI Project
OCR's Risk Analysis Initiative is the enforcement context every AI project now runs inside. Its 12th action, in March 2026, resolved a breach affecting 15 million individuals at a software company, and the initiative targets one failure, the lack of an accurate, thorough risk analysis of where ePHI lives and moves. An AI deployment adds new systems, new data flows, and new vendors, which means it either updates your risk analysis or invalidates it.
The financial stakes sit on top of the regulatory ones. IBM's 2026 Cost of a Data Breach report, as summarized by Becker's, puts healthcare breaches at an average of $6.64 million, down 10.5% from $7.42 million in 2025 and still the costliest of any industry for the 13th consecutive year. Those figures argue for doing the risk analysis at the diagnostic stage, while controls still cost only design effort.
State law adds a second compliance layer. California and Texas have healthcare AI rules in force, with disclosure duties when AI drafts clinical patient messages or is used in treatment, and most of what they cover sits on the clinical side. Colorado replaced its AI Act in May 2026 with a narrower law on automated decision-making that takes effect January 1, 2027 and still lists healthcare among its covered domains.
A consulting firm working across states should track these as part of the project, since HIPAA compliance alone no longer closes the question. The Security Rule itself is also in motion, with the final amended rule now projected for July 2027. The current rule stays in force until then.
---
How a HIPAA-Compliant AI Consulting Runs, Step by Step
Compliance holds when it is built into the delivery sequence itself, and the walkthrough below shows where each control lands in Boldr AI's delivery sequence.
The diagnostic phase opens with the BAA, executed before discovery touches any system containing PHI. The risk analysis is scoped into the diagnostic itself, mapping where PHI lives, which workflows move it, and what the planned automation changes, so the deliverable doubles as the documentation OCR expects to exist. PHI data-flow mapping then sets the build's boundaries, naming which systems the automation may read, write, and log.
In the build phase, workflows do the mechanical work and sensitive information stays with people wherever the process allows, with any PHI that does flow confined to BAA-covered services. Audit logging is designed into the system from the start, writing to the client's records so the client owns its own evidence. Before go-live, the deployment passes a compliance review against the risk analysis, and adoption training covers the human layer, including shadow-AI rules that keep staff from pasting patient data into consumer tools the project just made redundant.
After go-live, the operate phase keeps the risk analysis current as the system extends to new workflows. That cadence matches how OCR actually evaluates organizations, since a risk analysis is judged on being accurate and up to date, and an AI system that keeps changing needs documentation that keeps pace.
---
What HIPAA-Compliant AI Consulting Costs
Published pricing is rare in this market, and the honest guidance is shapes and ranges. Petronella publishes pricing that starts at $5,000 to $15,000 for a single-location practice and runs $25,000 to $75,000 for initial consulting and architecture at multi-location organizations, which is a fair market reference for standalone compliance assessment work. Full consulting projects structure differently, with a discovery phase first, deployment quoted on the scope that phase defines, and an operating team after go-live.
At Boldr AI, the Value Discovery Sprint is a one-time engagement from US$7,500 for US work. It opens with the BAA and carries the risk analysis scope and the PHI data-flow map. The Composed POD, which operates the system after go-live, starts from US$5,000/month.
Compliance work front-loads the diagnostic, and that is where a compliant project carries more work than a generic one. The risk analysis, data-flow mapping, and BAA chain setup happen before any build, which buys down both regulatory exposure and rework. A quote with a thin diagnostic and a fast build has left that work out, and the exposure it was meant to buy down stays with you.
---
Choosing a HIPAA-Compliant AI Consulting Partner That Survives the Audit
Trust is the actual product in regulated work, and in healthcare operations it is earned in the unglamorous places, meaning the risk analysis that was current, the log that captured what the agent did, and the BAA signed before discovery began. A medical group or community hospital choosing an AI partner is choosing whose discipline it will be audited on. Boldr AI structures healthcare work so the compliance evidence exists as a byproduct of how the work runs, and its Value Discovery Sprint starts provider organizations with the workflow diagnostic and the PHI data-flow map in a single scoped phase.
---
Frequently Asked Questions
Does our AI consulting firm really need to sign a HIPAA BAA?
Yes, if its work touches PHI at any point, including discovery and testing. Boldr AI executes the BAA before discovery begins, since workflow analysis in scheduling, intake, or revenue cycle systems exposes PHI almost immediately.
Can we use ChatGPT with patient data, or do we need a HIPAA-compliant AI consulting firm to do it?
Not on consumer or standard API tiers, which carry no BAA. PHI belongs only on BAA-eligible enterprise tiers of the major providers, configured so client data never trains the base model. Boldr AI builds on those tiers with training disabled.
What triggers OCR scrutiny on an AI consulting project?
Breaches and complaints, and the first document requested is the risk analysis. An AI deployment missing from it is a finding, which is why the update belongs in the diagnostic before go-live. Boldr AI scopes that update into its diagnostic.
Do state AI laws apply to healthcare operations projects?
Increasingly, yes. Colorado, California, and Texas are all reaching for healthcare AI, with more states drafting legislation, so HIPAA alone no longer answers the question. Boldr AI confirms current state-specific requirements at the start of each project.
How much does HIPAA-compliant AI consulting cost?
The Value Discovery Sprint is a one-time engagement from US$7,500 for US work. Deployment is quoted on the scope the sprint defines, and the Composed POD, which operates the system after go-live, starts from US$5,000/month. Boldr AI states both prices up front so the compliance work is budgeted before any build.
Schedule a Consultation
Boldr AI is an AI consulting and execution partner that fixes the process first, then automates what pays back, in systems the client owns.