- Date
- October 2006
- Location
- United States
- Industry
- Water & Wastewater
- Access vector
- External Remote Services
- OT targeting
- None
- Primary peril
- None
- Impact duration
- No outage
Harrisburg water filtering plant malware intrusion, 2006

The incident path, stage by stage
Each stage of the attack, from intent to financial loss, with this incident’s path highlighted. Solid cells are confirmed, dashed cells are reported by a source but not confirmed, dotted cells are implied by the facts, and Unknown marks what is not publicly known. Click a stage title or a highlighted cell to read that part of the analysis.
Scroll sideways to see every stage.
On this page
Highlights
What happened. In early October 2006, an employee’s laptop was compromised over the internet and then used as an entry point to install a virus and spyware on the computer system of a water filtering plant in Harrisburg, Pennsylvania[1]. The FBI did not believe the intruders were after the water system, and nothing happened to the water[1][2].
Why it matters. The sources disagree about how close the infected machine was to the treatment process, and the plant’s own answer has never been published[2][1]. It is an early example of commodity malware arriving through an ordinary IT path and landing somewhere nobody can later describe.
About this analysis. This analysis sets out what is known, stage by stage. The plant is not named in any source, and the public record consists of press reports from 2006 and a 2007 federal report that summarizes them. This account reflects the best available information as of October 22, 2026; new information could confirm, change or discredit parts of it.
Highlights
- An employee’s laptop, compromised over the internet, was used to install a virus and spyware on the plant’s computer system. No malware family has been named[1].
- The FBI believed the intruders wanted the computer for distributing e-mails or other electronic information, not to compromise the water system[1].
- The US Government Accountability Office (GAO) says the attack did not seem to target the control system directly, but that the software was capable of affecting water treatment. No source reports any effect on treatment[2].
Attack flow
- Updated: 2026-10-22
- Intent: Malicious [confirmed]
- Threat actors: Not attributed [unknown | “A foreign hacker”, unnamed]
- Access vector: External Remote Services [reported | T1133, employee remote access]
- IT impact: Malware [reported | Virus and spyware, T1105]
- OT targeting: None [confirmed | No OT consequence]
- OT sophistication: Not applicable [not in path | Commodity malware only]
- OT impact: None [not in path | Nothing done to treatment]
- Perils: No Perils [confirmed | No physical event]
- Financial losses: Cyber Incident Response [implied]
What happened
The intrusion took place in early October 2006, according to Computerworld; GAO dates it to October 2006[1][2]. It became public at the end of the month through press reporting, first by ABC News and then by UPI and, on November 1, Computerworld[1].
The path was ordinary. An employee’s laptop was compromised over the internet, and it was then “used as an entry point to install a computer virus and spyware on the plant’s computer system,” Computerworld reported, crediting ABC News[1]. A UPI report, which could not be re-checked for this analysis, and the RISI incident database, an aggregator, both say the employee had remote access, which the intruder used[3]. No official statement names the route, and no source names the malware.
FBI Special Agent Jerri Williams of the Philadelphia office told Computerworld: “We did not believe that they were doing it to compromise the actual water system, but just to use the computer as a resource for distributing e-mails or whatever electronic information they had planned.” Computerworld adds, paraphrasing her, that the FBI was still concerned the software could have interfered with the plant’s operations. No arrests had been made as of November 1[1]. According to RISI, the plant eliminated remote access and changed all passwords[3].
Intent: malicious
The intrusion was deliberate: someone took over an employee’s laptop and used it to install a virus and spyware[1]. What they wanted, in the FBI’s view, had nothing to do with water. Williams said the FBI believed they meant to use the computer “as a resource for distributing e-mails or whatever electronic information they had planned”[1]. That is the FBI’s reading of motive. No source says e-mail or any other content was actually sent.
Threat actors: not attributed
No one has been identified. GAO records “a foreign hacker,” drawing on press reporting[2], and Computerworld reports that the attackers were believed to have been operating outside the United States and that no arrests had been made[1]. No nationality, group or individual is established in any source, and no arrest was reported later. “A foreign hacker” is not an attribution, and no threat actor is highlighted on the map.
Access vector: an employee’s laptop and remote access
The entry point was an employee’s laptop compromised over the internet[1]. How the laptop itself was compromised is not described. UPI and the RISI database say the employee had remote access, which the intruder used, and the remedy reported afterward was to eliminate that remote access outright[3]. No official statement confirms the route, so External Remote Services is shown as reported.
The condition is a familiar one: remote access into a plant network, ending on a general-purpose machine that an employee also used on the open internet.
IT impact: a virus and spyware
A virus and spyware were installed on the plant’s computer system, according to Computerworld, and GAO says the intruder planted malicious software[1][2]. Both rest on press reporting rather than an official finding, so Malware is shown as reported. No source names the malware family, says how the files got onto the machine, or says what the spyware collected.
The FBI’s view that the intruders wanted to use the computer to distribute e-mails would amount to hijacking its resources, but no source reports that this happened[1].
OT targeting: no consequence, and a disputed machine
Nothing happened to the water treatment process, but the sources describe the infected machine differently. Computerworld calls it “the plant’s computer system” and describes it no further[1]. GAO says “the infection occurred through the Internet and did not seem to be an attack that directly targeted the control system,” while also saying the intruder planted software “capable of affecting the plant’s water treatment operations”[2]. UPI, in the report that could not be re-checked, described it as a computer that controls an important infrastructure system. That is the strongest claim and the weakest supported: it paraphrases the same FBI agent who told Computerworld the water system was not the target.
This analysis follows GAO: capability asserted, consequence not. GAO’s account is itself a summary of press reporting, and its footnote cites an ISA news item for this case[2]. The stage is shown as None, with no OT consequence.
OT sophistication: not applicable
There is no OT technique to grade. The intruders used a virus and spyware delivered through a compromised laptop[1], and nothing in the incident required knowing what a water plant is. The FBI believed the computer was wanted for its resources, not for its place in the plant[1]. The stage is marked not in path.
OT impact: nothing done to treatment
No source reports any change to water treatment. The line that gave the story its afterlife came from an industry expert, Mike Snyder of the Pennsylvania section of the American Water Works Association: “It’s a serious situation because they could possibly raise the level of chlorine being injected into the water”[1]. That was a general remark about what would be serious, not an account of this incident. No source alleges that anything was done to chlorine dosing, setpoints or process data. The stage is marked not in path.
Perils: none
There was no physical event. No source reports any change to water treatment, any equipment damage or any release; GAO, Computerworld and the FBI describe only malware on a plant computer[2][1]. No boil-water advisory or public-health effect appears in any source. The map shows No Perils.
The question that matters after a compromise is what the infected machine could reach, and here nobody has answered it.
Financial losses: none disclosed
No cost, outage duration or affected-customer count is disclosed, and because the plant is not named, even the population it serves is unknown. The response is implied rather than costed: the FBI’s Philadelphia office investigated[1], and the plant reportedly changed all system passwords and eliminated remote access[3]. Removing remote access is a standing fix rather than incident recovery, and it too carries no disclosed cost.
Loss evidence
- Implied: Network event response and recovery (password changes, FBI investigation); Voluntary improvements (removing employee remote access)
- Not supported: Direct business interruption (no treatment or service disruption reported); Third-party and societal impact (no boil-water advisory or public-health effect); Production data integrity (no process data or setpoints altered)
What this incident shows
Harrisburg is an early, clear instance of a pattern that still dominates OT incident reporting: commodity malware, arriving through an ordinary IT path, landing on a machine whose relationship to the process nobody can afterward describe[1][2]. The FBI thought the intruders wanted the computer for e-mail. GAO thought the software could have affected treatment. Both may be right.
The gap is the lesson. Nobody, not the plant, the FBI or GAO, states clearly what that computer could reach[2][1]. A plant that cannot answer that question after the fact has not documented the separation between its business and control systems, whatever separation it had in practice.
References
- McMillan, Robert. “Hackers break into water system network.” Computerworld / IDG News Service, 1 Nov. 2006. computerworld.com/article/1482209/hackers-break-into-water-system-network.html
- U.S. Government Accountability Office. “Critical Infrastructure Protection: Multiple Efforts to Secure Control Systems Are Under Way, but Challenges Remain.” GAO-07-1036, Sept. 2007, pp. 16-17. gao.gov/assets/gao-07-1036.pdf
- exida LLC. “Pennsylvania Water Company Hack.” RISI Online Incident Database, undated. risidata.com/Database/Detail/pennsylvania_water_company_hack


