The proposed HIPAA Security Rule update (90 FR 898), explained
HHS proposed on January 6, 2025 to replace the Security Rule’s flexible specifications with named controls, written inventories and fixed testing intervals. As checked on November 14, 2026, it is still a proposal, listed as a long-term action with final action shown for July 2027.
Where the proposal stands
The HIPAA Security Rule sets the safeguards that covered entities and business associates must apply to electronic protected health information. HHS first published it in 2003 and last revised it in 2013 [1]. On January 6, 2025, HHS published a notice of proposed rulemaking, “HIPAA Security Rule To Strengthen the Cybersecurity of Electronic Protected Health Information” (90 FR 898, RIN 0945-AA22), with comments due by March 7, 2025 [1] [2].
As checked on November 14, 2026, the proposal had not been finalized. The Federal Register lists one document under its regulation identifier, the proposed rule itself [3]. Its entry in the 2026 Unified Agenda, published August 14, 2026 [5], places the rule at the “Long-Term Actions” stage, marks it economically significant and major, and shows a final action date of July 2027 [4]. A long-term entry is a plan, not a promise: the text could change, slip or be withdrawn. Until a final rule appears, the current Security Rule, as last revised in 2013, stays in force.
HHS said in the proposal that it believed most of the existing rule’s obligations would not be substantially changed. Its stated aim was to write into the regulation, as explicit requirements, activities it already considered necessary, such as an inventory of technology assets as part of an accurate and thorough risk analysis [1].
One set of rules instead of required and addressable
Today the rule divides implementation specifications into “required” and “addressable”. The proposal would remove that distinction, so that a regulated entity would have to comply with both the standards and the implementation specifications, with specific exceptions written into some provisions, such as encryption [1]. Much of the proposal follows from that change: controls that an organization could previously document as not reasonable and appropriate would become requirements.
Written records of what an organization runs
The proposal opens the administrative safeguards with a new standard for a technology asset inventory. Each regulated entity would keep a written inventory of its technology assets, recording each asset’s identification, version, accountable person and location, and a network map showing how electronic protected health information moves through its electronic information systems [1]. Both would be reviewed and updated at least once every 12 months and whenever a change in the environment or operations may affect that information [1].
The risk analysis would become an explicitly written assessment with minimum contents, starting from a review of the inventory and network map to find where the information is created, received, maintained or transmitted [1]. A further new standard would require a documented compliance audit at least once every 12 months, covering each standard and implementation specification in the Security Rule [1].
Controls the proposal would name
The technical safeguards would move from general objectives to named controls. The main ones in the proposed text are these:
- Encryption: encrypt all electronic protected health information at rest and in transit, using encryption that meets prevailing cryptographic standards, with documented exceptions, such as a technology asset that does not support encryption [1].
- Multi-factor authentication: deploy it for all technology assets in relevant electronic information systems, defined as at least two of three categories of factor: something the user knows, has or is [1].
- Network segmentation: written policies and procedures ensuring that relevant systems are segmented to limit access to the information to authorized workstations [1].
- Malware and configuration: anti-malware protection for all technology assets in relevant systems, removal of extraneous software, and secure configuration of operating systems and software [1].
- Vulnerability scanning: automated scans at least once every six months, or more often if the risk analysis calls for it [1].
- Penetration testing: at least once every 12 months, or more often if the risk analysis calls for it, by a person with appropriate knowledge of and experience with generally accepted cybersecurity principles [1].
- Patching: within 15 calendar days for a critical risk and 30 calendar days for a high risk, counted from identifying the need where a patch is available, or from its release where it is not; other patches on timelines the entity sets in its own policies [1].
Recovery in 72 hours, notice in 24
The contingency planning standard would gain a written disaster recovery plan with procedures to restore critical relevant electronic information systems and data within 72 hours of a loss, and to restore other systems in the order set by a criticality analysis [1].
Two new notice clocks would run at 24 hours. A business associate would report to the covered entity the activation of its contingency plan without unreasonable delay and no later than 24 hours after activation. And where a workforce member’s access to another regulated entity’s systems changes or ends, the entity making the change would notify the other entity within 24 hours [1].
Business associates would have to show their work
Under the current rule, a covered entity obtains satisfactory assurances from a business associate through a contract. The proposal would add verification. At least once every 12 months, the covered entity would obtain written verification that the business associate has deployed the required technical safeguards, made up of two parts: a written analysis of the business associate’s relevant systems by a person with appropriate cybersecurity knowledge and experience, and a written certification, signed by a person with authority to act for the business associate, that the analysis was performed and is accurate [1]. A business associate would in turn obtain the same verification from its subcontractors [1].
The risk analysis would also include an assessment of the risks of entering into or continuing each business associate arrangement, based on that written verification [1]. For a health system with hundreds of vendors that touch patient data, that would make vendor oversight an annual evidence exercise rather than a contract clause.
Cost and timing if it is finalized
HHS estimated first-year costs of about $9 billion for regulated entities and affected group health plan sponsors together, and about $6 billion a year in years two through five for recurring compliance activities [1].
The proposal says a final rule would take effect 60 days after publication. Regulated entities would then generally have 180 days to comply with new or modified standards under 45 CFR 160.105, the statutory minimum, unless the final rule sets a different period for a provision [1]. The proposal also includes transition provisions for business associate agreements already in place [1].
Reading the proposal while it waits
Because the proposal is not law, none of its intervals or clocks bind anyone today. It still tells a security team where HHS thinks the current rule’s risk analysis and safeguards are headed, and several items, such as an asset inventory and a map of data flows, are things HHS already treats as part of an accurate and thorough risk analysis [1]. Organizations deciding how far to act before a final rule should weigh that against the possibility of changes, and confirm with counsel what the current rule requires of them.
The companion checklist on the HIPAA risk analysis under the rule as it stands sets out the steps the current text requires. Moraph’s teams help organizations with work like this; see our solutions.
Spotted an error? Tell us and we will check it and correct the explainer where needed.