Search
SEC News

Oracle Health Breach Hits 19.9 Million, Texas AG Says [2026]

Rachel Kowalski
Rachel KowalskiCybersecurity Analyst
13 min read
Oracle Health Breach Hits 19.9 Million, Texas AG Says [2026]

A cyberattack that Oracle quietly disclosed to hospital customers back in March 2025 has grown into one of the largest healthcare data breaches on record. According to a report filed with the Texas Attorney General and first detailed by Bloomberg on October 5, 2026, the breach of Oracle Health’s legacy Cerner systems compromised the personal and medical information of roughly 19.9 million people nationwide, with about 3 million of them Texas residents.

Oracle itself has not issued a public statement confirming that nationwide figure. The number comes from state breach-notification filings, the mechanism several state attorneys general use to track incidents affecting their residents, and it has climbed as more states report in. Fox News and CNET both cited the Texas filing when the story broke, and the tally has not moved much since, though Oracle has stayed quiet on specifics.

Google · Preferred Sources

Don't miss new tech stories on Google

Add TrendinTech once in the Google app and our stories appear in your news suggestions.

Add Now

What Happened to Oracle Health’s Cerner Systems

Oracle Health is the healthcare technology division Oracle built around its 2022 acquisition of Cerner, the electronic health record (EHR) giant that serves thousands of hospitals, clinics, and health systems across the United States. Not every Cerner customer runs on the same infrastructure. Years after the acquisition, a portion of Cerner’s older, legacy servers still held patient data that had not been migrated to Oracle’s own cloud platform. That older environment is where the intrusion reportedly took place.

Per filings reviewed by Becker’s Hospital Review, the breach began on January 22, 2025, was discovered on February 20, 2025, and the unauthorized access continued until April 1, 2025, a window of more than two months in which attackers had some form of access before the intrusion was fully shut down. Oracle began notifying affected healthcare customers in March 2025, meaning hospitals and clinics knew about the problem while the incident was technically still active.

That gap between discovery and containment echoes a pattern seen across the industry. Aggregate research from IBM and the Ponemon Institute found that healthcare data breaches took an average of 279 days to identify and contain in 2025, one of the longest detection windows of any sector they studied. Against that backdrop, Oracle’s roughly one-month gap between discovery and full containment looks relatively fast, even though the damage from the initial exposure was already done.

The Scale: 19.9 Million People, 3 Million in Texas Alone

The headline number traces back to a single document: a breach notification filed with the Texas Attorney General’s office. Under Texas law, companies that experience a breach affecting 250 or more Texas residents must file a notice with the state, and those filings become public record. Becker’s Hospital Review, which obtained the filing through a public records request, reported the Oracle Health breach affected 19.9 million people in total, with about 3 million of those being Texas residents, a figure Bloomberg and Yahoo Finance independently confirmed by reviewing the same filing.

To put that scale in context, 19.9 million people is roughly equal to the combined population of New York City and Los Angeles. It is also comparable in size to other major healthcare and retail breaches reported this year. Burger King’s Japan unit disclosed a breach affecting 11.18 million records in 2026, as this site reported in its coverage of the Burger King Japan data breach, while Denmark’s national breach involving reused “123456” passwords hit 8.8 million people, detailed in this site’s report on the Denmark password breach. Oracle Health’s number dwarfs both.

What Data Was Exposed

According to Bloomberg’s review of the Texas filing, the exposed information included Social Security numbers, addresses, and medical information. Some affected healthcare providers told patients directly that exposed records could include names, treating doctors, diagnoses, prescribed medicines, and test results, a combination that security researchers generally consider far more dangerous than a typical financial-data breach, since medical details cannot be changed the way a credit card number can.

That mix of identity and clinical data is exactly what makes healthcare breaches so costly to clean up. IBM and the Ponemon Institute put the average cost of a healthcare data breach at $7.42 million per incident in 2025, the highest average of any industry the firms studied. For comparison, the same research found the average cost of a data breach across all industries globally was $4.44 million in 2025, itself a 9 percent decrease from the prior year. Healthcare breaches, in other words, run well above the cross-industry average, and a breach of Oracle Health’s size sits at the far end of that curve.

Why Oracle Hasn’t Confirmed the Number

Oracle has not made a public statement confirming the nationwide total of affected individuals, according to SecurityWeek’s reporting on the story. That silence is notable given the scale involved, though it is not unusual for a company facing a breach of this size to let state filings do the talking while it manages notifications to individual healthcare customers directly. Oracle also has not disclosed a specific number of affected electronic health records, as distinct from affected people, leaving open questions about exactly how many patient charts, prescriptions, or test results were touched.

This pattern, where state attorneys general filings surface details that companies themselves have not volunteered, has become a recurring feature of how large breaches come to light. The UK’s Legal Aid Agency breach followed a similar arc, where the scope kept widening as more records came to light, a dynamic this site covered in its report on the UK Legal Aid Agency hack, where the exposed records dated back decades. Oracle Health’s situation differs in that the data is more recent, but the drip-feed disclosure pattern looks familiar.

Legacy Infrastructure Is the Real Story

The technical root of the Oracle Health breach is not exotic. It is a story about old servers. As BankInfoSecurity reported, the compromised systems were older Cerner infrastructure that had not yet been migrated to Oracle Cloud, meaning the data sat on hardware and software stacks Oracle inherited through the 2022 acquisition rather than infrastructure it built and hardened from scratch. Migration projects of that scale, moving decades of hospital records from one vendor’s systems to another’s cloud, routinely take years, and attackers have repeatedly shown they know how to find the gaps that open up during that transition.

That same theme, attackers exploiting the seams between old and new infrastructure, showed up elsewhere in 2026. Ivanti’s Connect Secure product suffered a zero-day that let attackers chain multiple bugs together to gain root access, a case this site covered when reporting on the Ivanti Connect Secure zero-day. Danfoss faced a similar problem when a vulnerability sat unpatched for 63 days before a related attack hit French retailers, detailed in this site’s coverage of the Danfoss vulnerability gap. In every one of these cases, the exploited weakness was not a brand-new system. It was an older piece of infrastructure that had outlived its planned lifespan without being fully retired or secured.

How This Compares to Other 2026 Healthcare and Retail Breaches

Becker’s Hospital Review tracked five healthcare vendor breaches in 2026 each affecting more than 1 million people, with their reported totals adding up to more than 30 million people combined, Oracle Health’s incident leading that list at nearly 20 million. The table below places Oracle Health’s breach next to other large-scale breaches reported on this site in 2026, to give a sense of relative scale.

BreachPeople AffectedData ExposedDisclosed By
Oracle Health (Cerner legacy systems)19.9 millionSocial Security numbers, addresses, medical informationTexas Attorney General filing
Burger King Japan11.18 millionCustomer recordsCompany disclosure
Denmark national breach8.8 millionReused “123456” passwords, personal dataGovernment notice
UK Legal Aid Agency250,000 (records dating to 1978)Legal aid applicant recordsAgency disclosure
Ledger hardware implant incidentNot fully disclosed; losses estimated at $93 millionCryptocurrency wallet accessCompany statement

Measured purely by people affected, Oracle Health’s breach is the largest healthcare-sector incident reported so far in 2026, and one of the largest breaches of any kind this site has covered this year. The comparison also shows how disclosure channels differ: some breaches come out through direct company statements, while others, like Oracle Health’s and the UK Legal Aid Agency’s, surface primarily through government filings and public records requests rather than voluntary corporate announcements.

The Regulatory Angle: Why Texas Matters

Texas’s breach notification law is one of the more aggressive state-level frameworks in the country, requiring disclosure to the Attorney General’s office once a breach affects 250 or more state residents, and making those filings subject to public records requests. That is precisely the mechanism that surfaced the Oracle Health figures. Without that filing, the scale of the breach might have stayed known only to the specific hospitals and health systems Oracle notified directly back in March 2025, each one potentially reporting a far smaller, hospital-specific number to its own patients.

Under HIPAA, any breach affecting 500 or more individuals must also be reported to the U.S. Department of Health and Human Services’ Office for Civil Rights, which maintains a public breach portal. Whether Oracle Health’s full 19.9 million figure appears there as a single combined entry or gets split across the dozens of hospital systems it serves will shape how visible the incident becomes to patients searching for their own exposure status. Fragmented reporting across many covered entities is common in EHR vendor breaches, and it is part of why the true scope of incidents like this one can take months, sometimes years, to fully surface.

Market and Industry Impact

Oracle’s stock did not see a dramatic single-day move tied directly to the October 5 Bloomberg report, consistent with how markets have generally treated healthcare data breaches at large, diversified technology companies in recent years: as a reputational and legal cost center rather than an existential threat to the business. Oracle Health represents one division within a much larger enterprise software and cloud company, and the financial exposure from the breach, likely to include regulatory fines, patient notification costs, credit monitoring offers, and potential class-action litigation, will play out over years rather than in a single quarter’s earnings.

That said, the reputational stakes for Oracle Health specifically are real. The division competes directly with Epic Systems and other major EHR vendors for hospital contracts, and breach history is increasingly a factor hospital IT and compliance teams weigh during vendor selection and renewal cycles. A breach of this size, tied explicitly to legacy infrastructure Oracle inherited through acquisition, gives competitors a talking point and gives existing Cerner customers a reason to ask pointed questions about migration timelines for their own data.

Historical Context: A Pattern of Healthcare Vendor Breaches

Healthcare has been one of the most heavily targeted sectors for cyberattacks for years, and breaches tied to third-party vendors and legacy systems are not new. What makes the Oracle Health case distinctive is the combination of scale and the clear acquisition-related cause: data sitting on servers that were in the process of being retired or migrated, rather than data on systems a company built and secured from day one. That is a structural risk every large tech acquisition in healthcare carries, and it is one regulators and hospital customers are likely to scrutinize more closely going forward.

Becker’s Hospital Review’s broader tracking shows this is not an isolated event. Five separate healthcare vendor breaches in 2026 alone affected more than 1 million people each, and their combined totals exceed 30 million, with Oracle Health accounting for nearly two-thirds of that combined figure on its own. That concentration in a single vendor’s incident, rather than spread evenly across many smaller breaches, is itself a signal of how much patient data rides on a small number of EHR platforms.

What Patients Should Do Now

Patients who received care through a hospital or clinic using Cerner systems during the January to April 2025 window should watch for a notification letter from their specific healthcare provider, since Oracle’s own notifications went out to individual hospital and health system customers rather than directly to every patient. Anyone concerned about exposure, particularly given that Social Security numbers were reportedly included, should consider placing a fraud alert or credit freeze with the major credit bureaus and monitoring statements for unfamiliar medical billing, a known tactic in medical identity theft where stolen insurance information gets used to bill for services the patient never received.

Because login credential reuse remains one of the most common ways stolen personal data gets turned into further account compromise, affected individuals should also treat this as a prompt to review password hygiene across medical portals, insurance accounts, and financial accounts. This site’s guide to the best password managers for 2026 walks through setting up unique, randomly generated credentials for exactly this kind of scenario, where a single breach at a vendor can expose information tied to many downstream accounts.

Predictions: Where This Goes From Here

  • Expect additional state attorney general filings in the coming months as more states report their resident counts, likely pushing media coverage of the “nationwide total” higher or clarifying it with more precision than the current 19.9 million estimate.
  • Expect Oracle to face at least one class-action lawsuit tied to the breach, following the near-universal pattern after healthcare breaches of this scale in the United States.
  • Expect hospital customers to accelerate migration timelines off legacy Cerner infrastructure, using this breach as use in contract renewal and modernization budget conversations.
  • Expect regulators, including HHS’s Office for Civil Rights, to open a formal review of the incident given the scale, consistent with how HHS has handled past breaches affecting millions of patients.
  • Expect competitors in the EHR market to use the breach as a competitive argument in sales conversations with hospitals currently evaluating vendor options.

The Cost Math Behind Breaches This Size

Industry-wide breach cost data gives a rough sense of what incidents like this one can ultimately cost, even before litigation outcomes are known. The table below breaks down the IBM and Ponemon Institute figures most relevant to a breach of Oracle Health’s profile.

MetricFigureSource Year
Average cost of a healthcare data breach$7.42 million per incident2025
Average cost of a data breach, all industries$4.44 million (down 9% year over year)2025
Average time to identify and contain a healthcare breach279 days2025
Organizations reporting breaches involving AI systems13%2025

These figures, drawn from IBM’s annual Cost of a Data Breach research, represent averages across a single incident, not aggregate totals for a breach the size of Oracle Health’s. Given that 19.9 million people were affected, well above the typical sample size behind these averages, the real cost of notification, legal defense, credit monitoring, and potential regulatory penalties could run meaningfully higher than the per-incident average once all of it is tallied.

Why Legacy EHR Infrastructure Keeps Causing Problems

Electronic health record systems are notoriously difficult to modernize because they sit at the center of clinical workflows that cannot tolerate downtime. A hospital cannot simply take its EHR offline for a weekend migration the way a retailer might swap out a point-of-sale system overnight. That operational reality is part of why so much Cerner infrastructure remained on legacy hardware years after Oracle’s 2022 acquisition closed, and it is a dynamic that will likely keep surfacing in other large healthcare IT consolidations as well.

The parallel to other 2026 infrastructure-related security incidents is hard to miss. Whether it is Ivanti’s chained zero-day bugs, Danfoss’s 63-day unpatched exposure window, or Oracle Health’s un-migrated Cerner servers, the common thread is the same: attackers are consistently finding their way in through systems that organizations knew needed attention but had not yet gotten around to retiring or hardening.

Frequently Asked Questions

How many people were affected by the Oracle Health data breach?
According to a filing with the Texas Attorney General reported by Bloomberg and Becker’s Hospital Review, roughly 19.9 million people nationwide were affected, including about 3 million Texas residents. Oracle has not publicly confirmed this nationwide total itself.

When did the Oracle Health breach happen?
The breach reportedly began on January 22, 2025, was discovered on February 20, 2025, and the unauthorized access continued until April 1, 2025. Oracle began notifying some healthcare customers in March 2025.

What systems were affected?
The breach involved older Cerner servers whose patient data had not yet been migrated to Oracle Cloud. Cerner became part of Oracle following the company’s 2022 acquisition of the electronic health record vendor.

What data was exposed?
Reports citing the Texas filing say exposed information included Social Security numbers, addresses, and medical information. Some affected healthcare providers told patients records could include names, treating doctors, diagnoses, prescribed medicines, and test results.

Has Oracle confirmed the 19.9 million figure?
No. Oracle has not made a public statement confirming the nationwide total of affected individuals, according to SecurityWeek. The figure comes from a state breach-notification filing rather than a direct Oracle announcement.

How does this compare to other 2026 data breaches?
At nearly 20 million people, the Oracle Health breach is larger than other major 2026 incidents such as the Burger King Japan breach (11.18 million records) and Denmark’s national password breach (8.8 million people), making it one of the largest healthcare data breaches disclosed so far this year.

What should affected patients do?
Patients who received care through a hospital using Cerner systems during the January to April 2025 window should watch for a direct notification from their provider, consider a credit freeze or fraud alert given that Social Security numbers were reportedly exposed, and review password hygiene across medical and financial accounts.

Could this lead to regulatory penalties for Oracle?
Breaches affecting 500 or more individuals must be reported to the U.S. Department of Health and Human Services’ Office for Civil Rights under HIPAA, and breaches of this scale have historically drawn regulatory review and, in many prior cases, settlements or corrective action plans.

Related Coverage

Rachel Kowalski

Rachel Kowalski

Cybersecurity Analyst

Rachel Kowalski is the Cybersecurity Analyst at TrendinTech, where she covers vulnerabilities, ransomware, supply chain attacks, nation-state operations and the security rules shaping organizations in the United States and Europe. She writes to make risks and countermeasures understandable to people who are not specialists.

All stories by Rachel Kowalski (254)

Related Articles