Election security help came on request in 2018, which meant it could be withdrawn
For the 2018 midterms, DHS offered election offices free, voluntary services only when they asked, while responsibility stayed local. That left the durability of each kind of help to federal choices, a gap that opened when a key federal agreement ended in 2025.
Looking Back revisits past weeks one at a time, up to the most recent, and asks what it meant for the banks, health systems and public agencies we serve. The series was written in 2026, with the benefit of hindsight, and each article says which week it looks back on.
Two-minute summaryThe dek and the first sentence of each section, quoted from the article
For the 2018 midterms, DHS offered election offices free, voluntary services only when they asked, while responsibility stayed local. That left the durability of each kind of help to federal choices, a gap that opened when a key federal agreement ended in 2025.
- The week On Monday, November 5, 2018, as the midterm elections arrived, the Department of Homeland Security, the Department of Justice, the Office of the Director of National Intelligence and the FBI issued a joint statement on Election Day preparations.
- Information: alerts, councils and the sharing centers The first kind of support was knowledge about threats.
- Assessment: testing systems before an election The second kind of support was assessment.
- Monitoring: sensors and continuous watch The third kind of support was monitoring.
- Response: advisors and help when something breaks The fourth kind of support was response.
- Objections: the help was free, the federal role grew, the risk is overstated The first objection is that worrying about durability is ungrateful.
- With hindsight The model outlasted 2018.
The week
On Monday, November 5, 2018, as the midterm elections arrived, the Department of Homeland Security, the Department of Justice, the Office of the Director of National Intelligence and the FBI issued a joint statement on Election Day preparations. It was released in the names of Secretary Kirstjen Nielsen, Attorney General Jeff Sessions, Director of National Intelligence Dan Coats and FBI Director Christopher Wray.
The statement said the agencies had been working to combat influence efforts and to support state and local officials in securing elections, including efforts to harden election infrastructure against interference. It said that, at that time, the agencies had no indication of compromise of the nation’s election infrastructure that would prevent voting, change vote counts or disrupt the ability to tally votes. It referred to an October 19 joint statement on foreign interference and to an executive order signed on September 12 that, in the agencies’ words, made clear the government would not hesitate to defend electoral processes or punish those who tried to undermine them. It said the agencies had been preparing for nearly two years and were closely engaged with officials on the ground, and that they would stay focused on the mission long after polls closed. It closed with advice to voters: get election information from state or local election offices, and think before sharing election-related information, asking who wrote it, who posted it and what their sources were.
The arrangement behind the statement was set out on the DHS election security page, in the version captured by the Internet Archive on November 6, 2018. Election infrastructure had been designated in January 2017 as a subsector of the Government Facilities critical infrastructure sector. Under that designation, DHS, through its National Protection and Programs Directorate (NPPD), offered services that state and local election officials could use to reduce cyber and physical risk, on a prioritized basis at their request.
The page divided responsibilities plainly. State and local election officials managed and secured election infrastructure day to day, administered the statewide voter registration databases required by the Help America Vote Act, trained local officials, tested and certified voting equipment, and disbursed funds for equipment and administration. DHS shared threat information, provided cybersecurity assistance by request, fostered coordination between the election community and DHS, and helped develop and deploy sensors to detect malicious activity, again by request.
The page described the services in one sentence: they were available at no cost to state and local governments and officials, were available on request and were strictly voluntary, and NPPD provided them only when requested by state and local election officials. The services listed were regional Cybersecurity Advisors and Protective Security Advisors; assessments such as Cyber Hygiene scanning, Risk and Vulnerability Assessments and Cyber Resilience Reviews; detection and prevention such as threat hunting and Enhanced Cyber Services; information sharing through national alerts and the Homeland Security Information Network; around-the-clock incident response; and training. The page also described a Government Coordinating Council of election officials from all levels and a Sector Coordinating Council of private sector owners and operators, each organized under the national critical infrastructure partnership model.
The November 2018 midterms were held under the critical infrastructure designation DHS had made for election infrastructure in January 2017. The joint statement on the eve of the vote presented federal support as a coordinated national effort. The DHS page behind it described something more specific: a menu of free services, delivered only when a state or local office asked, with responsibility for running and securing elections left where it had always been. On the eve of the vote, the agencies reported no indication of a compromise that would prevent voting or change counts. Its weakness was structural, and it showed later. Support that arrives on request, at no cost, through federal budgets and agreements can be reduced or withdrawn by federal decisions, while the duty to run a secure election stays local.
This article argues that an election office, or any state or local agency that relies on federal cyber support, should classify each form of that support by how durable it is, and plan for the less durable forms to change. The 2018 menu falls into four kinds of support, and they divide cleanly. The first is information: threat alerts, coordination councils and the information sharing centers. The second is assessment: scans and reviews that test an office’s systems at a point in time. The third is monitoring: sensors and threat hunting that watch networks continuously. The fourth is response: advisors and incident response when something goes wrong. Each depends on a different federal mechanism and leaves a different gap when it changes. A final section takes the objections.
The categories are exclusive in what they do for an office. Information tells it what threats exist. Assessment tells it what weaknesses it has. Monitoring tells it what is happening now. Response helps it when an attack succeeds. Together they cover the cycle of security work, and an office can be well served in one and exposed in another.
For each kind, the same three questions decide durability. Who delivers it: federal staff, a state team, a nonprofit under a federal agreement, or a vendor? Who pays: a federal appropriation, a cooperative agreement, a grant or the office’s own budget? And what remains with the office if the arrangement ends: knowledge, fixed weaknesses, data, relationships, or nothing? The answers differ across the four kinds, and they can differ for the same kind of support from one office to the next.
The argument reaches beyond elections. Counties, cities, school districts, public hospitals and utilities draw on the same federal programs. Election offices are a sharp case because their deadlines are fixed in law and their failures are public, but the classification applies to any public body. Where a statute or grant condition imposes specific obligations, agencies should confirm them with their own counsel.
The 2017 designation is the starting point. Designating election infrastructure as critical infrastructure left the running and securing of elections with state and local officials. The 2018 page described what it did instead: it let DHS provide services on a prioritized basis at the request of state and local officials, and it brought the structure of the national infrastructure protection partnership, with its coordinating councils, to the election community. Every element of federal support therefore rested on two things outside the election office’s control, the federal decision to offer a service and the federal priority given to the office’s request. Those two dependencies are the root of the durability question.
The joint statement’s key sentence also shows the limits of federal visibility. The agencies said they had no indication of compromise that would prevent voting, change counts or disrupt tallying. Federal agencies could see what their own intelligence, their sensors and the reports of state and local partners showed them. An office that had not requested a sensor, had not joined an information sharing center and did not report an incident may have been largely outside that view. Federal assurance at the national level was therefore built from local participation, and its reach depended on the same voluntary choices as the services.
Information: alerts, councils and the sharing centers
The first kind of support was knowledge about threats. The 2018 page listed national alerts and advisories and access to the Homeland Security Information Network. It described the Government Coordinating Council, through which election officials from local, state and federal government shared information and coordinated, and the Sector Coordinating Council of private sector owners and operators. The joint statement itself was a form of this support: a public assurance, backed by intelligence agencies, about what was and was not known.
Information is the cheapest support to give and the easiest to underrate. A county election office with a small staff cannot run its own threat intelligence. It depends on others to tell it which phishing campaigns target election officials, which vulnerabilities in common systems are being exploited, and which foreign activity is under way. In 2018 that knowledge reached offices through several channels at once: federal alerts, the coordinating councils, state election directors and the information sharing and analysis centers that served state and local governments.
Not every channel was federal
Some of those channels were run by federal agencies. Others were run by nonprofit organizations funded in part by federal agreements. An office that received an alert did not always know which. That distinction later mattered. When a federal agreement changes, the channel it funds may change too, while a channel run directly by federal staff may continue. An office that knows the source of each information feed can see which ones depend on a funding decision outside its control.
What information support cannot do
Information support has a ceiling. An alert is useful only if someone reads it, understands it and can act. A small office receiving national alerts may lack the staff to decide which apply to its systems. The value of information support therefore depends on the office’s own capacity, and on intermediaries, such as a state election office or a state security team, that translate national alerts into local action. When we think about durability, the intermediary matters as much as the source.
Building a local information baseline
An office that wants information support to survive changes in any one channel can build a baseline of its own. That means subscribing to the free federal alerts directly, joining the coordinating structures available to it, maintaining a relationship with the state’s security team, and keeping a list of its key system vendors with contacts for security notices. Each of these stands apart from any single funding agreement. Together they ensure that the loss of one channel reduces the office’s information rather than ending it.
The office as a source of information
The joint statement ended by telling voters to get election information straight from the source, their state or local election office. That advice turned election offices into an information channel in their own right, and a target for impersonation and false claims. An office’s website, social media accounts and phone lines are part of its security posture. Protecting them means securing the accounts with strong authentication, registering look-alike domains where practical, and having a plan to correct false information quickly through channels voters already trust. This work depends on no federal service, and it is where an office’s own effort has the most direct effect on public confidence.
Measuring information support
An office can measure whether its information support works. Useful questions include how many alerts it received in the last quarter that applied to its systems, how long it took to act on them, and whether any incident it experienced had been described in an alert it missed. A state can ask the same questions of its counties. The answers show whether information is reaching people who can act, and which channels deliver it, which is the evidence needed when one channel changes.
States as intermediaries
Between national channels and a county office sits the state. A state election office, a state chief information security officer or a state fusion center can receive national information, filter it for the systems the state’s counties use and push it to the people who can act. Where that intermediary works, a county’s information support depends less on any single national channel. Where it does not, each county must do the filtering itself. States that want their counties to be resilient to changes in national programs can invest first in that intermediary role, which is cheaper than replacing the national channels and serves every county at once.
Assessment: testing systems before an election
The second kind of support was assessment. The 2018 menu listed Cyber Hygiene scanning, Risk and Vulnerability Assessments and Cyber Resilience Reviews. These differ. Scanning checks internet-facing systems for known vulnerabilities on a recurring basis. A risk and vulnerability assessment is a deeper, hands-on test, including attempts to exploit weaknesses. A resilience review examines an organization’s practices against a framework. All three were offered at no cost, by request.
Assessment is valuable because it brings expertise an office cannot afford to employ. A county that has never had a penetration test can learn, from one federal assessment, that its voter registration system’s administrative interface is exposed or that its staff accounts lack multifactor authentication. The findings are durable even if the service is not: a fixed weakness stays fixed.
Assessments are a queue
Services on request at no cost can create demand that exceeds supply. The 2018 page said services were provided on a prioritized basis. In practice that means an office may wait, and an office that asks late in an election cycle may not be served before the election. Assessment support is therefore most useful when an office plans it into its calendar, well ahead of the periods when its systems matter most, and treats the federal assessment as one input in a recurring cycle rather than a one-time event.
Owning the findings
A federal assessment produces a report. The report belongs to the office, and so do the fixes. An office should track every finding to closure with an owner and a date, and retest when it can. A finding left open because the office assumed the federal team would follow up is the most avoidable kind of exposure. For state election offices that coordinate many counties, aggregating findings by type across counties shows where a statewide fix, such as a shared configuration standard or a common contract term, would close several gaps at once.
Replacing assessment if it becomes scarce
If federal assessment capacity shrinks, an office needs another source. Options include state security teams, university partners and commercial testers paid through state or federal grants. The planning question is not whether a substitute is as good as the federal team, but whether the office has a written plan for its next assessment that does not depend on a single provider’s availability.
What each assessment answers
The three 2018 services answer different questions, and an office should know which it needs. Recurring scanning answers whether internet-facing systems carry known, exploitable weaknesses this week. A hands-on risk and vulnerability assessment answers whether an attacker with skill and time could get in and how far they could go. A resilience review answers whether the office’s practices, such as backups, access management and incident planning, are organized to withstand and recover from an attack. An office that has had only scanning knows little about its practices. One that has had only a resilience review knows little about its exposed systems. A planned cycle uses each for its purpose.
Vendors inside the assessment scope
Election offices rely on vendors for voting systems, electronic poll books, results reporting and hosting. CISA’s 2026 plan notes, for example, that electronic poll books used to update voter records or register new voters will be connected to the internet and should be secured, including multifactor authentication for administrative functions. An assessment that stops at the office’s own network may miss the systems that matter most. Offices should ask vendors for their own assessment results, include vendor-managed connections in scope where contracts allow, and write assessment cooperation into new contracts.
Timing assessments against the certification cycle
The 2018 page listed testing and certifying voting equipment among the states’ responsibilities. Certification fixes a system’s configuration for use. An assessment that finds a weakness in certified equipment may not be fixable before an election without recertification. Offices and states should therefore schedule hands-on assessments of the systems around certified equipment, such as networks, management workstations and reporting systems, where fixes are within the office’s control, and should raise findings about certified equipment through the certification process rather than through ad hoc changes.
Findings as a budget argument
An assessment report is also evidence for a budget request. A county finance committee may be skeptical of an abstract request for security funding. It is harder to dismiss a federal report listing specific exposed systems and the cost of fixing them. Offices should use findings that way, and should keep the history of findings and fixes across cycles, which shows a governing body that money spent on security produced measurable change.
Monitoring: sensors and continuous watch
The third kind of support was monitoring. The 2018 page said DHS would help develop and deploy sensors to detect malicious activity, by request, and listed threat hunting and Enhanced Cyber Services under detection and prevention. Monitoring differs from assessment in time. An assessment is a snapshot. Monitoring watches network traffic and alerts continuously, so that an intrusion between assessments can be seen.
Continuous monitoring is the support a small office is least able to provide for itself. It requires sensors, analysts reading alerts around the clock and the experience to tell an attack from noise. Delivered as a service, it can give a county the same watch as a large state. That is also why it creates the deepest dependence. An office that relies on an outside monitoring service may have no internal capability at all, so the loss of the service leaves it blind rather than less sighted.
The mechanism behind the sensor
Monitoring services for state and local governments can be delivered through more than one arrangement: directly by federal teams, through federally funded nonprofit centers, through state programs and through commercial providers. The question for an office is which arrangement its own sensor sits under, who reads the alerts, under what agreement, and what happens to the sensor and the alerts if that agreement ends. By 2025, CISA was telling state and local partners that used Albert sensors to coordinate directly with the Center for Internet Security for that service, which showed that the arrangement mattered.
Data and access when a service changes
A monitoring service holds data about the office’s network: traffic records, alerts, investigation notes. When a service changes hands or ends, the office needs to know whether it can retrieve that history, and whether the provider retains access to the office’s network. These are contract questions, and for free services there may be no contract the office negotiated. An office should at least know the terms it accepted and should ask, while the service is running, how an exit would work.
Building partial independence
A small office cannot replace a monitoring service with its own staff. It can reduce its dependence by keeping its own logs for a defined period, enabling alerts in the cloud and email services it already pays for, and arranging with its state or a regional partner for a backup watch. Partial independence means that the loss of the primary service degrades monitoring instead of ending it.
What monitoring needs to see
An election office’s network is not uniform. CISA’s 2026 plan lists the components: voter registration systems, election management systems, tabulation systems, results reporting, public information systems, electronic poll books, and the wider state and local networks they connect to, including email. A sensor at the county’s internet connection sees traffic to and from the outside. It may not see activity inside a hosted voter registration system run by the state or a vendor, or on a results reporting site hosted elsewhere. An office should map which components each monitoring arrangement covers and which fall to someone else, and should know who is watching each one on election night.
Who reads the alert in the middle of the night
A sensor produces alerts. Someone has to read them, decide which matter and call the office. In a shared monitoring service, that someone works for the provider, and the call goes to whatever contact the office registered. An office should test that path: confirm the registered contact is current, that the contact can be reached outside business hours during election periods, and that the person reached has authority to act. A monitoring service whose alerts reach an outdated email address is no service at all on the night it matters.
Monitoring and voter data
Monitoring an election network means an outside party may see traffic involving voter data. Offices should know what data the monitoring arrangement collects, where it is kept, who can see it and for how long, and whether those terms meet state law on voter records. Those questions are easier to settle when a service is set up than when it is changing hands, and their answers become part of the exit plan if the arrangement ends.
Watching the vendors’ way in
Vendors that support election systems may connect to them remotely for maintenance, updates or ballot preparation. Those connections are a path into the office that the office did not build and may not see clearly. Monitoring should include them: which vendor accounts exist, when they connect, from where and what they do. An office that cannot answer those questions from its own records depends on the vendor’s honesty and the monitoring service’s coverage at once. Requiring vendors to connect only through an office-controlled path, with logging the office keeps, reduces both dependencies.
Response: advisors and help when something breaks
The fourth kind of support was response. The 2018 menu offered regional Cybersecurity Advisors and Protective Security Advisors who provided immediate and sustained assistance, coordination and outreach, and incident response with around-the-clock intrusion analysis. Response support is the one an office hopes never to need and the one it most needs to have arranged in advance.
For an election office, the timing of response is unforgiving. An incident in the days before an election, or while results are being counted and certified, has to be contained without delaying deadlines fixed in law. An office that first looks for help at that moment loses time it does not have.
Advisors as relationships
Regional advisors are people. Their value lies in knowing the office before an incident: its systems, its staff, its state partners. That relationship is durable only as long as the advisor’s position and the office’s contact with it continue. An office should know its advisors by name, should involve them in exercises, and should have a second contact if the first changes role.
Writing response into the election calendar
An election office’s incident response plan should map onto its election calendar. It should say who decides, during early voting, on election day and during canvass, whether to take a system offline; who contacts state officials and federal partners; and how paper records and manual procedures substitute for affected systems. The 2020 statement by the election infrastructure coordinating councils later emphasized that paper records of each vote allowed recounts where results were close. Paper and manual procedures are the office’s own response capability, and they do not depend on any outside service.
Arranging backup responders
Federal incident response capacity is finite and is prioritized. A large incident affecting several jurisdictions at once may stretch it. An office should therefore have a second responder arranged in advance, through its state, a regional partner or a contracted provider, with the contracts, approvals and access details ready. In a crisis, procurement rules do not pause.
Physical security is part of response
The 2018 menu paired Cybersecurity Advisors with Protective Security Advisors, and the 2026 plan lists voting locations, storage facilities, processing facilities and administrative offices as election infrastructure. Response planning for an election office covers threats to people and places as well as networks: a threat against a counting center, a disruption at a ballot printing contractor, a storage facility breach. Physical incidents also have fixed deadlines. The advisors who help assess facilities before an election are the natural contacts when something goes wrong, which is another reason to know them before it does.
Exercises that use the plan
Response plans are tested by exercises. CISA publishes election security tabletop exercise packages with template scenarios and discussion questions. An office that runs one each cycle, with its state partners, its vendors and its federal advisors in the room, finds out whether its contacts are current and whether its decision rules work under time pressure. The exercise is also a chance to rehearse the fallback: what happens if a federal partner or an outside service is unavailable that day.
Recovery against the certification deadline
After an incident, recovery has its own clock. Canvass and certification deadlines are set in law, and systems taken offline must be restored, or replaced by manual processes, in time to meet them. Recovery plans should say which systems are needed for canvass, how they would be rebuilt from known-good configurations, and who verifies them before use. Backups held where an attacker who compromised the office’s network cannot reach them, tested before each election period, are the foundation. They are also entirely within the office’s control.
Keeping the record of what happened
Every incident and exercise should end with a written record: what happened, what worked, which outside services were used and how quickly they responded. Over several cycles, those records show which dependencies proved reliable under pressure. They are also the institutional memory that survives staff turnover, which in small offices can remove most of an office’s experience in one departure.
Speaking during an incident
Response includes communication. When an election system is affected, voters, candidates and the press will ask what happened before the facts are known. An office should prepare holding statements, agree in advance with state officials who speaks for whom, and decide what it will say about the integrity of results while an investigation runs. Federal partners can help with analysis. They cannot speak for a county about its election. That voice is local, and so is the responsibility for getting it right.
Objections: the help was free, the federal role grew, the risk is overstated
The first objection is that worrying about durability is ungrateful. The 2018 services were free, and offices that asked for them received expertise they could not otherwise afford. The point stands. The argument here is not that offices should have refused the help. It is that free help on request creates a dependence whose terms the office does not set. Using the help while planning for its change is prudent, not ungrateful.
The objection that the federal role kept growing
A second objection is that the federal role did not shrink overall. After 2018, election infrastructure stayed designated as critical infrastructure, CISA continued to describe election security as one of its highest priorities, and in September 2026 it published an election infrastructure security plan describing no-cost, voluntary services. The record bears this out, and it supports the argument. The services run directly by federal staff, such as assessments and advisors, persisted in the 2026 plan. The changes came in the arrangements run through outside organizations under federal agreements. Durability varied by mechanism, which is the classification this article recommends.
The objection that grants replaced services
A third objection is that the 2025 model did not withdraw support but changed its form, offering grant funding through the State and Local Cybersecurity Grant Program alongside no-cost tools and advisors. Money that an office controls may even be more durable than a service it requests. That depends on the grant. Grants require applications, administration and procurement, which small offices may struggle to manage, and grant programs have their own funding cycles set by Congress. Converting a free service into a grant-funded purchase moves the work of sourcing, contracting and managing the service to the office. The classification still applies: an office should know which of its services now depend on a grant, and what happens when the grant period ends.
The objection that federal involvement should be smaller
A fourth objection runs the other way: elections are run by states, and federal involvement in them should be limited. This article does not argue for more federal authority. The 2018 model was voluntary, and its 2026 successor remains so. The argument is about planning within that model. Because the federal role is voluntary on both sides, an office cannot assume any particular service will be there next cycle, and should plan as if each one might change.
The objection that this is a budget question
A fifth objection is that the issue is money, not security: if the funding is there, services continue, and if it is not, planning cannot conjure it. Planning cannot create money, but it changes what money buys and when. An office that knows which services might change can ask for funding before the change rather than after, can share costs with its state or neighbors, and can decide which services matter enough to pay for. An office that discovers its dependence when a service ends faces the same choices with less time and less bargaining power over price.
The objection that the risk was overstated
A sixth objection is that the risk was overstated: on the eve of the 2018 vote the agencies had no indication of a compromise that would prevent voting or change counts, and in 2020 the coordinating councils found no evidence that any voting system had been compromised. Security programs are judged by what does not happen, so this is weak evidence that the risk was overstated. It is stronger evidence that the combination of local responsibility, federal support and paper-based resilience worked in those years. The argument is about keeping that combination working when one part changes.
The objection that small offices cannot do more
A final objection is that small election offices, some with a few staff, cannot classify dependencies or arrange backups. Their states can. A state election office or state security team can do the classification for all its counties, negotiate statewide backup arrangements and fund them through state appropriations or federal grants. The classification is light work for a state and impossible for some counties, which is why states are the natural owners of it.
What we would tell a state election director today
List every outside service your counties use for information, assessment, monitoring and response. For each, record who provides it, under what agreement, who pays, and what happens to its data if it ends. Mark the ones that depend on a federal agreement with an outside organization. For those, arrange a fallback before the next election cycle, funded and contracted. Review the list each year when the state budget is set, so that any change in a federal program is matched by a decision in the state’s own budget rather than discovered in the middle of an election. And keep the parts that depend on no one, paper records and manual procedures, tested and ready.
What we would tell a county or city IT director
The same exercise applies beyond elections. A county that uses federal scanning for its public website, a regional sensor for its network and a state team for incident response has the same four kinds of dependence. Map them, note which rest on agreements the county did not sign, and decide which ones the county would pay to replace if they ended. Put those decisions in the budget request before the change, not after it. Where the county’s election office shares its network, make sure the election office’s dependencies are on the same list, so that a change in one service does not surprise both.
With hindsight
The model outlasted 2018. On November 12, 2020, the executive committee of the Election Infrastructure Government Coordinating Council and the members of the Sector Coordinating Council, including CISA, issued a joint statement calling the November 3 election the most secure in American history. They said all states with close results in the presidential race had paper records of each vote, and that there was no evidence any voting system deleted or lost votes, changed votes or was in any way compromised. They pointed to pre-election testing and state and federal certification of voting equipment as further sources of confidence.
The mechanisms changed later. On September 29, 2025, CISA announced that its cooperative agreement with the Center for Internet Security would reach its planned end on September 30, 2025. It said it was moving to a model supporting state, local, tribal and territorial governments through grant funding under the State and Local Cybersecurity Grant Program and the Tribal Cybersecurity Grant Program, no-cost services and tools such as Cyber Hygiene scanning, the Cybersecurity Performance Goals, regional advisors, professional services and regular operations calls. It said it would continue to collaborate with the Multi-State Information Sharing and Analysis Center, and that partners using Albert sensors should coordinate directly with CIS for that service.
The September 2025 release still listed Cyber Hygiene scanning among CISA’s no-cost services, the same service named on the 2018 page. CIS describes the consequences on its own pages. Its MS-ISAC membership FAQ says the move to fee-based membership responded to federal funding cuts that culminated in the end of federal support beyond September 30, 2025, that CIS had temporarily funded MS-ISAC services since March 2025, and that organizations not taking paid membership would begin to lose access after October 1, 2025 in a phased process. It says election offices that join receive an elections threat intelligence package. On another page CIS states that, due to the termination of funding by DHS, it no longer supports the Elections Infrastructure ISAC.
Federal staff services continued. CISA’s 2026 Election Infrastructure Security Plan, dated September 2026, says Secretary Mullin tasked CISA in July with developing it. It says that, at the request of state and local officials, CISA provides no-cost, voluntary cyber and physical security services, and that state and local election officials have primary responsibility for protecting election infrastructure, across more than 10,000 local jurisdictions. It lists components to protect, from voter registration systems to electronic poll books, and says poll books connected to the internet should use multifactor authentication for administrative functions. CISA’s election security page says its services are available at no cost to state and local officials and private sector election infrastructure partners.
What held up was the 2018 division of responsibility, restated in CISA’s 2026 plan. What changed was the funding of information sharing and monitoring run through an outside organization. As checked on November 14, 2026, offices that had relied on those channels without a fallback had to pay for membership or find another route, while the services federal staff delivered directly were still on offer.
Spotted an error? Tell us and we will check it and correct the article where needed.