Congress bought the threat-sharing half of the 2015 cyber package and left breach notice to the states

In January 2015 the White House asked government to play four cyber roles: clearinghouse, breach referee, prosecutor and privacy guardian. Only the clearinghouse passed in anything like the proposed form, and agencies still work inside that split.

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

In January 2015 the White House asked government to play four cyber roles: clearinghouse, breach referee, prosecutor and privacy guardian. Only the clearinghouse passed in anything like the proposed form, and agencies still work inside that split.

  1. The week On Monday, January 12, 2015, President Obama went to the Federal Trade Commission in Washington and announced a set of consumer privacy proposals (PBS).
  2. Clearinghouse: DHS as the hub for threat data The sharing proposal was modest in mechanism.
  3. Breach referee: one national notice standard The breach proposal asked for much more.
  4. Prosecutor: new tools against cybercrime The criminal-law proposals were a mixed set (White House, January 13).
  5. Privacy guardian: rules for the government’s own handling The fourth role was the least visible.
  6. Where the four-role reading is weakest One objection is that the four roles were never meant to stand alone.
  7. What agency leaders can do with this history The split of 2015 has practical consequences for public agencies and for the regulated organizations that deal with them.
  8. With hindsight Congress acted on the clearinghouse role within the year.

The week

On Monday, January 12, 2015, President Obama went to the Federal Trade Commission in Washington and announced a set of consumer privacy proposals (PBS). The next day he went to the Department of Homeland Security’s National Cybersecurity and Communications Integration Center, the NCCIC, in Arlington, Virginia, and announced a cybersecurity legislative proposal (White House, January 13). The announcements came in the run-up to the State of the Union address on January 20.

The Monday package centered on a bill the White House called the Personal Data Notification & Protection Act. The fact sheet said it would set a single, national standard and a 30-day deadline, from discovery of a breach, for companies to notify customers. It would also criminalize illegal overseas trade in Americans’ identities (White House, January 12). The fact sheet linked the proposal to breaches that had exposed the personal and financial information of tens of millions of Americans.

The same fact sheet carried a mix of other items. JPMorgan Chase and Bank of America, working with FICO, would offer free credit scores to card customers, as would USAA, State Employees’ Credit Union and Ally Financial for their customers. The administration backed the FTC’s IdentityTheft.gov site for victims. It proposed a Student Digital Privacy Act limiting the use of data collected in schools. It also said the Commerce Department would release revised Consumer Privacy Bill of Rights legislation within 45 days (White House, January 12).

Law firms that read the breach bill text that week described its mechanics in more detail. Notice would go to individuals within 30 days, normally by mail or phone. Firms could skip notice if a risk assessment found no reasonable risk of harm, but would have to send the assessment to the FTC. The FTC would enforce the law as an unfair or deceptive practice, state attorneys general could also sue, and there would be no private right of action. Most state breach laws would be preempted (Mintz).

The Tuesday package had three parts (White House, January 13). The first encouraged private companies to share cyber threat information with the NCCIC. The center would pass it to other federal agencies and to private Information Sharing and Analysis Organizations, or ISAOs, as close to real time as practicable. Companies that shared would get targeted liability protection, provided they removed unnecessary personal information. DHS and the Attorney General, consulting the Privacy and Civil Liberties Oversight Board, would write guidelines for how the government received, kept, used and disclosed the data.

The second part updated criminal law. It would allow prosecution of the sale of botnets, criminalize the overseas sale of stolen U.S. financial information, expand authority to deter the sale of spyware, and give courts power to shut down botnets used in attacks. It would extend the Racketeer Influenced and Corrupt Organizations Act to cybercrime and modernize the Computer Fraud and Abuse Act. The fact sheet said the CFAA changes would put insignificant conduct outside the law while making clear that insiders who abuse their access for personal purposes could be prosecuted (White House, January 13).

The third part restated the breach bill announced the day before. The fact sheet said it would replace the patchwork of 46 state notification laws, plus those of the District of Columbia and several territories, with one federal statute (White House, January 13). It also announced a White House Summit on Cybersecurity and Consumer Protection at Stanford University on February 13. Separately, it announced $25 million over five years from the Department of Energy for a cybersecurity education consortium of 13 historically Black colleges and universities and two national laboratories.

The setting was tense. Sony Pictures was still recovering from the attack the government had attributed to North Korea in December. On the same Monday the President spoke at the FTC, attackers claiming loyalty to the Islamic State took over U.S. Central Command’s Twitter and YouTube accounts for a short time. Congress had passed several narrower cyber laws in December 2014, including an update to the Federal Information Security Management Act for federal agencies. The broad proposals of January 2015 aimed at what those laws had left out.

The January 2015 proposals are usually remembered as a single push for cyber legislation that partly succeeded. We read them as four separate job offers to government, each with a different future. Congress accepted one, half-accepted two and declined one. Agencies and the organizations they regulate still work inside that split.

The four roles map onto the two fact sheets of that week. Together they cover everything the package asked government to do about cyber risk:

  • Clearinghouse: DHS, through the NCCIC, would collect threat information from companies, pass it to other agencies and private groups, and give sharers liability protection.
  • Breach referee: the FTC would enforce one national breach notification standard with a 30-day clock, replacing most state laws.
  • Prosecutor: federal prosecutors would get new and clarified criminal tools, from botnet takedowns to a revised Computer Fraud and Abuse Act.
  • Privacy guardian: DHS and the Attorney General would write and enforce rules on how the government handled the personal information that came with shared data.

We take each role in turn, then the strongest objections to reading the package this way, and then what an agency program director or chief information security officer can do with the result. The argument is that the clearinghouse passed because it asked little of anyone, while the referee failed because it asked states to give up power. That difference explains a great deal of the compliance work public agencies and regulated firms still carry.

One clarification on scope. The breach bill applied to businesses, not to federal agencies. Federal agencies’ own security and breach duties came from a different statute, the Federal Information Security Modernization Act of 2014, and from guidance by the Office of Management and Budget. State and local agencies answer mainly to their own state laws. The January proposals touched public bodies in a different way: as the hub for shared data, as enforcers, and as regulators whose authority might be preempted.

Clearinghouse: DHS as the hub for threat data

The sharing proposal was modest in mechanism. It did not order anyone to share. It offered companies liability protection if they shared cyber threat indicators with the NCCIC and met privacy conditions, and it encouraged private ISAOs to form for sharing among companies (White House, January 13). The NCCIC would then distribute what it received to relevant federal agencies and to the ISAOs.

What the hub actually needed

A clearinghouse has three jobs, and each is an operational build for the agency that runs it. It must accept data in a consistent format, at machine speed, from many sources. It must decide quickly what to pass on and to whom, stripping anything it should not hold. And it must send data out in a form recipients can use without manual work.

In January 2015 the NCCIC already operated as a 24-hour center, and DHS had been working on machine-readable formats for indicators. The proposal assumed DHS could scale that work into a national service. That assumption put a large program on one agency’s operations budget, while the benefits accrued mostly to the companies and agencies that received the data.

Why liability protection mattered so much

Companies were reluctant to share for reasons that had little to do with technology. Their lawyers worried that sharing an indicator might breach a customer contract, a privacy law or antitrust rules, or that it might be used against them later in litigation or by a regulator. Liability protection addressed those fears directly. It made sharing safer for the sharer without making it mandatory.

That design choice is why the clearinghouse role attracted broad support. It cost companies nothing they did not choose to spend. It gave DHS a central role without new regulatory power. It gave privacy advocates a hook, through the personal information conditions, to argue for narrower terms. Everyone could find something to like.

What it meant for state and local government

State and local agencies had a double role. They were potential recipients of federal threat data, often through the Multi-State Information Sharing and Analysis Center and state fusion centers. They were also potential sharers, since their own networks saw attacks first. The proposal encouraged both flows but funded neither. A county information technology office with a small security staff could join a sharing group. It could not easily act on a high-volume indicator feed without tools to ingest it.

The practical test for any public body was whether shared data changed its defenses. An indicator that arrives by email and is read by a person the next morning does little. An indicator that arrives by machine feed and updates a firewall block list within minutes does a lot. The clearinghouse model worked best for agencies that had built the receiving end.

The limits of a voluntary hub

A voluntary system shares what participants choose to share. Large financial firms and technology companies already traded detailed threat data through their own groups, such as the Financial Services Information Sharing and Analysis Center. For them, the government hub was one more source, useful mainly for what only the government could see. Smaller organizations had the most to gain but the least capacity to contribute or consume.

That asymmetry was visible from the start. The proposal’s value depended on two-way flow, yet companies feared giving too much and agencies were cautious about releasing classified or sensitive data. A hub that receives more than it returns loses participants. The design gave DHS the job of fixing that imbalance without new tools to do it.

The ISAO idea

Sector-based Information Sharing and Analysis Centers had existed since a 1998 presidential directive on critical infrastructure. They covered industries such as finance, energy and health. The January 2015 proposal used a broader term, Information Sharing and Analysis Organizations, to encourage groups that did not fit a sector (White House, January 13). A regional group of manufacturers, a set of cities, or companies sharing a supply chain could form one.

For government, the ISAO idea was a way to reach organizations that the sector centers missed. Many small and mid-sized bodies, including local governments, school districts and utilities, had no natural home for sharing. A looser model let them organize around geography or shared vendors. The cost was fragmentation. More groups meant more places for an indicator to stall, and more need for the hub to connect them.

For agency program directors, the practical question was which groups to join and how to avoid duplicating effort. An agency in several groups receives the same indicator several times in different formats. Without automation to deduplicate and prioritize, more sharing can mean more noise. The value lies in the groups whose members face the same attackers and can turn shared data into action.

Measuring whether the hub works

The proposal said little about how success would be measured. A clearinghouse can count indicators received and sent, but those numbers describe activity, not value. A better measure is whether recipients blocked attacks they would otherwise have missed, and how quickly. That requires feedback from recipients, which voluntary systems rarely collect.

For an agency joining such a system, the same logic applies inside its own walls. Track how many shared indicators were applied to controls, how many generated detections and how long each step took. Those numbers tell the agency whether its participation is worth the staff time. They also give it evidence when budgets are set.

Breach referee: one national notice standard

The breach proposal asked for much more. It would set one federal standard: notice to individuals within 30 days of discovering a breach of sensitive personal information (White House, January 12). The FTC would enforce it, state attorneys general could bring cases, and most state breach laws would be preempted (Mintz).

What the patchwork looked like

The fact sheet counted 46 state breach notification laws in January 2015, plus those of the District of Columbia and several territories (White House, January 13). Each set its own definition of personal information, its own trigger, its own timing language and its own rules on notifying the state. Some required notice to the attorney general. Some specified content. Many said only that notice should come without unreasonable delay.

A national company with a breach had to work through every applicable law at once, because affected individuals lived in many states. Lawyers built matrices to do it. The cost fell on the breached organization and, indirectly, on the people waiting for notice while the analysis ran.

The definitions problem

The hardest part of the patchwork was not timing but definition. Each state decided what counted as personal information. Most covered a name combined with a Social Security number, a driver’s license number, or a financial account number with the code needed to use it. Some added medical information, biometric data, online credentials or other identifiers. The same incident could be a notifiable breach in one state and not in another.

Definitions decide everything downstream. They decide whether notice is required, to whom and how fast. A national standard would have collapsed that analysis into one test. Without it, organizations have to classify their data against many tests, and keep the classification current as states amend their laws.

Public agencies faced the same work for their own data. A state agency that serves residents of other states, such as a university system, a tollway or a licensing board, may face questions about which state’s law applies. Counsel usually has to answer that before an incident, because the answer shapes logging, data retention and contracts.

What a 30-day clock demands

A 30-day outer limit sounds generous. In operation it is tight. Within that window an organization has to confirm a breach happened, establish what data was involved and whose, decide whether notice is required, prepare letters, set up a call center and print and mail notices. Each step depends on the one before. Forensic scoping often takes weeks on its own.

The White House proposal counted the 30 days from discovery (White House, January 12). That choice put a premium on knowing what data sat where. An organization with good data inventories and record-level logs can scope a breach in days. One without them may spend most of the window learning what it held. The clock, in effect, rewarded preparation more than response.

The proposal’s risk-assessment exemption and law enforcement delay offered some flexibility (Mintz). Both required documentation. An organization relying on either would need a written assessment and a record of the law enforcement request. Those records become the evidence an enforcer reviews after the fact.

What the referee role would have changed

For businesses, a single standard promised certainty. One clock, one definition, one regulator. The White House said so explicitly, describing the proposal as giving companies the certainty of a single, national standard (White House, January 12). For consumers, the 30-day deadline was a clear outer limit. A fixed outer limit can also become the default timetable, where a state rule of notice without unreasonable delay might push some firms to move faster.

For government, the referee role moved power. The FTC would gain rulemaking authority and become the central enforcer. State attorneys general would keep a role but subject to the FTC’s right to intervene. State legislatures would lose the ability to set stricter rules for most of the field (Mintz). That last change was the crux. Preemption was the feature businesses wanted most and the feature states liked least.

Why referees are harder to create than hubs

A clearinghouse adds a service. A referee replaces existing referees. Every state that had passed a breach law had made choices: what counts as personal information, whether encryption creates a safe harbor, whether to notify the state. Some states had gone further than the federal draft. Preemption asked them to give those choices up.

The risk-assessment exemption was a second sticking point. It let a company skip notice if it concluded there was no reasonable risk of harm, provided it told the FTC (Mintz). Consumer advocates saw it as a loophole. Industry saw it as necessary to avoid notices for trivial incidents. The same argument had already played out in many state legislatures, with different results in different states.

Where public agencies sat in this fight

Federal agencies were outside the bill’s scope. Their breach duties ran through the 2014 federal security law and OMB guidance. State agencies were regulated by their own state laws, many of which applied to government bodies as well as businesses. A national private-sector standard would have left those public duties in place while removing the parallel private ones. Agencies that ran shared services with private partners, such as health exchanges or payment processors, would have faced two regimes on either side of a contract.

The federal agencies’ parallel track

Federal agencies were moving on their own track that winter. The Federal Information Security Modernization Act of 2014, signed in December, required agencies to notify Congress of major incidents. It also directed the Office of Management and Budget to update its guidance on responding to breaches of personal information. Those duties applied whatever happened to the private-sector bill.

The result was an odd asymmetry. Had the January proposal passed, a company would have faced a 30-day statutory clock enforced by the FTC. A federal agency holding the same kind of data would still have followed OMB guidance and its own policies. Agency leaders who wanted consistency had to build it internally, by aligning agency breach procedures with the standards they expected of their contractors. One practical route was to write notice timelines and content rules into contracts with service providers, so that a contractor’s breach would be handled on terms the agency had chosen.

What the referee role asked of the FTC

The proposal would have expanded the FTC’s workload considerably. It would gain rulemaking authority over breach notice, receive risk assessments from companies claiming the exemption, and coordinate with the Federal Communications Commission and the Consumer Financial Protection Bureau where their jurisdictions overlapped (Mintz). That is a significant operating program for an agency whose privacy staff was small relative to the economy it covered. The proposal said little about resources, which is a common gap in legislation that assigns new enforcement duties.

State attorneys general were the public officials most directly affected. Their offices enforced the state breach laws and, in many states, received breach notices. A federal statute with FTC primacy would have reshaped their caseloads and their influence over national companies. Their position on preemption mattered to whether Congress would move.

The case states made for keeping their role

States had practical arguments beyond turf. Many had updated their laws after local incidents, and their attorneys general had experience enforcing them. Some states required notice to their own offices, which gave them early warning of breaches affecting residents. A federal standard enforced mainly by the FTC would centralize that information in Washington. States also argued that they could move faster than Congress when new types of data or new attacks emerged.

Those arguments carried weight because the federal draft was, in some respects, weaker than the strongest state laws. A preemptive federal law sets a ceiling as well as a floor. For a state that had already gone further, accepting the federal standard meant reducing protection for its residents.

Prosecutor: new tools against cybercrime

The criminal-law proposals were a mixed set (White House, January 13). Some targeted specific behaviors: selling botnets, selling stolen U.S. financial information overseas, selling stalking spyware. Some extended existing laws: RICO to cybercrime, clearer penalties aligned with similar non-cyber crimes. One gave courts authority to order the shutdown of botnets used for denial-of-service attacks and other crimes.

Botnet takedowns and the role of courts

The botnet proposal built on practice. The Justice Department had already used civil injunctions under a fraud statute to disrupt botnets, including the Coreflood botnet in 2011 and the Gameover Zeus botnet in 2014. Those actions depended on showing fraud. A botnet used mainly for denial-of-service attacks, rather than for stealing banking credentials, might fall outside that authority. The proposal would have closed the gap by letting courts order takedowns for a wider range of botnet crimes (White House, January 13).

For public agencies, takedowns matter in two ways. Agencies are frequent targets of denial-of-service attacks, especially around elections, tax deadlines and benefits launches. Agencies’ own devices can also be among the infected machines in a botnet. A takedown order that redirects infected machines to a government-controlled server raises questions about what that server collects, which links back to the privacy guardian role.

Stolen financial data sold overseas

Two proposals addressed the market for stolen data. One would criminalize the overseas sale of stolen U.S. financial information, such as card and bank account numbers. The Monday fact sheet added a crime for illegal overseas trade in identities (White House, January 12). Both targeted the stage after a breach, where criminals turn data into money.

These changes would have mattered most to prosecutors building cases against brokers of stolen data who never touched a victim’s network. For banks, card issuers and agencies that issue identity documents, the effect would be indirect: a higher cost for the criminal market that uses their data. Those organizations’ own controls would still have to do most of the work.

The Computer Fraud and Abuse Act question

The most contested piece was the CFAA. The fact sheet described two changes in tension. Insignificant conduct would fall outside the law, which answered critics who said the statute could criminalize breaking a website’s terms of use. At the same time, the law would clearly cover insiders who abuse their access for personal purposes (White House, January 13).

That second change mattered to every organization that holds sensitive data. An employee who uses legitimate access to look up a celebrity’s record, a neighbor’s tax file or a former partner’s address is an insider abusing access. Agencies and firms wanted a clear criminal backstop for that behavior. Civil liberties groups worried that tying criminal liability to an employer’s use policy would make ordinary policy violations into federal crimes.

Why this role was half-accepted

Pieces of the prosecutor role were adopted over the following years through different channels, some through rules of procedure rather than statute. The insider clarification in the CFAA was never enacted. As the hindsight section explains, the courts later moved the law in the opposite direction. The result is that organizations cannot rely on criminal law alone to deter insiders who stay within their technical permissions.

What it meant for agency programs

For agencies that hold sensitive records, such as tax, benefits, licensing and criminal justice, insider misuse is a standing risk. Their defenses have to come mainly from access design, logging and administrative discipline, with criminal referral as a last resort. The January 2015 proposal would have strengthened that last resort. Its failure leaves the first lines carrying more weight.

Program directors can make this concrete. Every system that shows personal records should log record-level access, review unusual patterns and show staff, at login, that access is monitored and limited to official duties. Those steps work whatever the criminal law says.

The point applies across levels of government. A federal benefits agency, a state tax department and a county sheriff’s records unit all hold data that employees could misuse for personal reasons. Each has its own disciplinary rules and, in many states, state computer crime laws that may reach conduct the federal statute does not. Program directors should know which of those apply to their staff and make sure policies, training and access banners are consistent with them.

Privacy guardian: rules for the government’s own handling

The fourth role was the least visible. The sharing proposal required companies to remove unnecessary personal information before sharing, as a condition of liability protection. It also required DHS and the Attorney General to write guidelines, after consulting the Privacy and Civil Liberties Oversight Board and others, on how the federal government would receive, keep, use and disclose the information (White House, January 13).

Why this role was needed

The proposal came less than two years after public disclosures about intelligence collection had made companies wary of handing data to the government. A sharing program without privacy rules would have faced both political resistance and reluctance from companies whose customers would object. The privacy guardian role was the price of the clearinghouse.

In operational terms, the role meant DHS had to build filters. Each incoming indicator had to be checked for personal information that was not needed to understand the threat. An email address used by an attacker might be relevant. A victim’s name usually was not. Getting that right at machine speed is hard, and errors run both ways: stripping too much makes the indicator useless, stripping too little breaks the promise.

Speed against minimization

The proposal asked for sharing as close to real time as practicable and for privacy review before sharing (White House, January 13). Those goals pull against each other. Human review of each indicator is slow. Automated review is fast but imperfect. The only way to have both is to agree in advance which fields an indicator may carry, and to design formats that leave personal information out by default.

That is a design decision, not a legal one, and it belongs to the agencies that build the systems. An agency that defines a narrow indicator schema can share fast with little risk. An agency that shares raw logs or full email messages will either share slowly or share too much. The 2015 debate treated privacy mainly as a legal condition. In operation it is mostly a data design problem.

Who checks the guardian

The proposal gave oversight a role through the Privacy and Civil Liberties Oversight Board and through published guidelines. For agencies receiving data from DHS, the practical consequence was a chain of handling rules: what they could keep, for how long and for what purposes. Each receiving agency needed its own procedures to honor those rules, which meant records schedules, access controls and training for the analysts who handled the data.

What it meant for state and local recipients

State fusion centers and local agencies that received federal threat data inherited the handling rules. That could be a burden for smaller offices without privacy staff. It also gave them a framework they often lacked for their own data. A county that learns how to strip personal information from indicators before sharing them is better placed to handle its own breach and records obligations.

Where the four-role reading is weakest

One objection is that the four roles were never meant to stand alone. The White House presented a package, and Congress traditionally takes packages apart. Reading the outcome as a set of accepted and rejected roles may impose more design on the result than the politics justify. That is fair as a description of how bills move. It does not change the consequences. The roles that passed and the roles that failed shaped compliance work differently, whatever the reasons.

A second objection is that the breach referee role did not fail so much as move. Federal regulators later wrote incident reporting rules for their own sectors, which many observers read as a partial national standard by other means. We take that point seriously and return to it in the hindsight section. But sector reporting to regulators is not the same as notice to individuals. The patchwork for individual notice remains, and it is the part that most affects people whose data is exposed.

A third objection is that the clearinghouse was less of a success than its passage suggests. Passing a law is not the same as building a working hub. Participation in federal sharing programs has been debated in oversight reports for years. We agree that passage overstated the result. The argument here concerns which roles became law, not whether each worked as hoped.

A fourth objection is that preemption was not the only obstacle. Industry groups disagreed among themselves over notice triggers, retailers and banks argued over who should pay for breaches, and the congressional calendar was crowded. All true. Preemption was still the issue that set states against the bill, and states had real influence over their delegations.

A fifth objection holds that the patchwork has benefits the article understates. States experiment, and good ideas spread. Several state laws added features, such as notice to attorneys general or coverage of online credentials, that a 2015 federal standard might have frozen out. We accept that the patchwork has produced useful innovation. The cost is borne by every organization that must comply with all of it at once, and by individuals whose notice depends on where they live.

What agency leaders can do with this history

The split of 2015 has practical consequences for public agencies and for the regulated organizations that deal with them. Each of the four roles points to a specific piece of work.

Build the clearinghouse connection on your side

An agency should know what threat data it shares, with whom, in what format and on what legal basis. It should also know what it receives and what happens to it. The useful measure is time from receipt of an indicator to action in the agency’s own controls. If the answer is days, the agency is a subscriber, not a participant. Automating that path is usually a modest project with a large effect.

Keep a live map of breach notice duties

Because the referee role failed, every organization that holds personal information needs its own map. For a state agency, that means its own state law, any federal program rules attached to grants or data, and contractual duties to partners. For a federal agency, it means OMB guidance and agency policy. The map should list each duty’s trigger, clock, recipient and owner, and be tested in exercises.

Do not rely on criminal law for insiders

Access design, record-level logging and administrative sanctions have to carry the weight that a clearer CFAA would have shared. Agencies should review their highest-risk record systems for those controls and document how misuse would be detected and handled.

Write down your own privacy rules for shared data

An agency that receives or shares threat data should have written rules on personal information in that data: what it strips, what it keeps, for how long and who can see it. Those rules protect the agency when a share later becomes controversial. They also speed up sharing, because the privacy decision has been made in advance rather than case by case.

For regulated organizations that deal with agencies

Banks, health systems and other regulated organizations face the same split from the other side. They share threat data with the federal hub under the 2015 law’s protections. They notify individuals under state laws. They report to their own regulators under sector rules. Each of those duties has its own clock and recipient. A single incident can set all of them running at once.

The practical step is to merge these into one incident playbook with a single decision log. Who decides that an event is a reportable incident, under which rule, by what time, and who sends the report. Regulators increasingly expect to see that log. Agencies that oversee these organizations can help by aligning their own reporting forms and timelines where the law allows.

The sharing law that eventually passed in 2015 has had a fragile life since 2025, as the hindsight section describes. Agencies and partners that rely on its protections should track its status as a recurring legal item, and know what they would do if the protections lapsed again.

In concrete terms, that means three things. Legal staff should know which sharing arrangements depend on the law’s protections and which rest on contracts or other authority. Security staff should know whether partners would pause sharing during a lapse, and what the agency would lose if they did. And leadership should decide in advance whether the agency would keep sharing on its own legal footing if the federal protections were gone for a time.

With hindsight

Congress acted on the clearinghouse role within the year. The Cybersecurity Information Sharing Act of 2015 passed as Title I of the Cybersecurity Act of 2015, enacted in December 2015 within that year’s omnibus spending law. It gave private entities liability protection for monitoring their own systems, taking defensive measures and sharing cyber threat indicators, along with antitrust protection and exemption from disclosure laws for shared information (CRS). DHS ran the federal intake, including the Automated Indicator Sharing program, which pushes indicators to participants by machine.

The law differed from the White House draft in ways that favored sharers, and the privacy conditions became a requirement to remove personal information not directly related to a threat. DHS and the Justice Department issued the privacy and handling guidelines the law required in 2016. In 2018 Congress reorganized DHS’s cyber work into the Cybersecurity and Infrastructure Security Agency, which took over the NCCIC’s functions.

The law was written to expire after ten years (CRS). Its protections lapsed on October 1, 2025, during a government shutdown. Congress restored them through January 30, 2026 in the stopgap funding law that ended the shutdown in November 2025, then extended them to September 30, 2026 in a February 2026 funding law. By the Congressional Research Service’s September 2026 update, Congress had extended them to December 11, 2026 (CRS). A long-term reauthorization had bipartisan support but had not passed by that update. For a law built to give companies legal certainty, the run of short extensions was a poor fit.

The referee role never arrived. No federal breach notification standard for private companies was enacted. States kept legislating instead, and by 2018 every state had a breach notification law. Many states later shortened their deadlines, widened their definitions of personal information or added notice to the attorney general. The patchwork the 2015 fact sheet wanted to replace grew denser.

Federal regulators built sector rules in the gap. The federal banking agencies adopted a rule in 2021 requiring banks to notify their primary regulator of significant computer-security incidents within 36 hours, effective in 2022. The SEC adopted rules in 2023 requiring public companies to disclose material cybersecurity incidents, generally within four business days of determining materiality. In 2024 it amended Regulation S-P to require customer notice within 30 days for the firms it covers. Congress passed the Cyber Incident Reporting for Critical Infrastructure Act in 2022, requiring covered entities to report significant incidents to CISA. The final rule was due by October 2025. CISA missed that date, and by autumn 2026 the final rule was still under review rather than in force. These are reporting duties to regulators or investors, mostly not notice to individuals.

The prosecutor role was adopted only in pieces. An amendment to Federal Rule of Criminal Procedure 41, effective December 2016, allowed judges to issue warrants for remote searches in some botnet and concealed-location cases. The insider clarification to the CFAA was not enacted. In June 2021 the Supreme Court decided Van Buren v. United States and read the statute’s “exceeds authorized access” clause narrowly, so that a person who misuses information they are entitled to reach does not violate that clause. In May 2022 the Justice Department announced a charging policy saying it would not charge good-faith security research under the CFAA. The law moved away from the insider coverage the 2015 package wanted.

The privacy guardian role survived inside the sharing law. Federal agencies’ own security came under heavy scrutiny within months of the January proposals. In June 2015 the Office of Personnel Management disclosed the first of two intrusions that together affected personnel and background-investigation records of more than 20 million people. That event did more to change federal security practice than any part of the January package.

What held up is the basic diagnosis. Voluntary sharing with legal protection could pass. A national notice standard that preempted the states could not. Eleven years later, agencies and regulated organizations still run threat sharing through a federal hub and still map breach duties state by state.

Spotted an error? Tell us and we will check it and correct the article where needed.

  • information sharing
  • breach notification
  • DHS
  • Computer Fraud and Abuse Act
  • state law
  • privacy