Altaba’s SEC order tied misleading filings and failed controls to one breach not properly assessed

The SEC’s $35 million order found Yahoo’s risk factors, financial discussion and sale agreement misleading, and its disclosure controls deficient, for one reason. The breach its security team found reached managers and lawyers but was not properly assessed for disclosure.

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

The SEC’s $35 million order found Yahoo’s risk factors, financial discussion and sale agreement misleading, and its disclosure controls deficient, for one reason. The breach its security team found reached managers and lawyers but was not properly assessed for disclosure.

  1. The week On Tuesday, April 24, 2018, the Securities and Exchange Commission announced that Altaba Inc., formerly Yahoo Inc., had agreed to pay a $35 million penalty.
  2. Risk factors: a risk described in the future tense after it happened Risk factors describe what could go wrong.
  3. Management’s discussion and analysis: known trends the filings omitted The MD&A requires management to discuss known trends, demands, commitments, events or uncertainties that are reasonably likely to affect liquidity or results.
  4. Transaction representations: what the buyer was told The third part concerns the sale of Yahoo’s operating business.
  5. Controls conclusions: certifications that the system worked The fourth part concerns the company’s statements about itself.
  6. Why the security finding stalled before disclosure Each of the four parts failed at the same junction: between the security team’s knowledge and the people responsible for a public statement.
  7. How the lesson applies across sectors The Altaba order applied securities law to a public company.
  8. What the order did not decide The order was a settled administrative proceeding.
  9. Objections: materiality, timing, uncertainty and process Four objections deserve a direct answer.
  10. What a leader could check The four parts translate into checks that need no new system.
  11. With hindsight The Altaba order was followed by further enforcement on the same point.

The week

On Tuesday, April 24, 2018, the Securities and Exchange Commission announced that Altaba Inc., formerly Yahoo Inc., had agreed to pay a $35 million penalty. The SEC said the company had misled investors by failing to disclose a data breach in which hackers stole personal data relating to hundreds of millions of user accounts. The company neither admitted nor denied the findings in the SEC’s order.

The order described the breach and what the company knew. In late 2014, Yahoo learned of a breach of its user database. By December 2014, its information security team, including its chief information security officer, had determined that hackers associated with the Russian Federation had stolen copies of database files. They held the personal data of at least 108 million users, and likely the entire user database. The order said the security team referred to the stolen data as the company’s “crown jewels”: usernames, email addresses, telephone numbers, dates of birth, hashed passwords, and security questions and answers.

According to the order, within days members of senior management and the legal team received internal reports from the chief information security officer that the theft had occurred. The order quoted Yahoo’s own later filing: the “relevant legal team had sufficient information to warrant substantial further inquiry in 2014, and they did not sufficiently pursue it.” The order found that senior management and legal staff did not properly assess the scope, business impact or legal implications of the breach. It also found they did not share information about it with the company’s auditors or outside counsel to assess disclosure obligations.

The order identified three kinds of statement that the undisclosed breach made misleading. The first was risk factors. In its annual reports for 2014 and 2015 and its quarterly reports for the first three quarters of 2015 and the first two of 2016, Yahoo described security breaches as a risk the company faced, without saying one had occurred. The second was the management’s discussion and analysis, which omitted known trends or uncertainties the breach presented. The third was a July 23, 2016 stock purchase agreement with Verizon, filed as an exhibit to a Form 8-K. In it, Yahoo represented that, to its knowledge, there had been no security breaches that could reasonably be expected to have a material adverse effect on the business.

The order also recounted a fourth kind of statement, which the company itself corrected: its conclusion, in each of those periodic reports, that its disclosure controls and procedures were effective. Its 2016 annual report later said those controls were not effective for the periods from December 31, 2014 through September 30, 2016. It attributed that “exclusively to deficiencies in the Company’s existing security incident response protocols related to the 2014 Security Incident.” The order also noted that the risk factors were incorporated into registration statements for employee stock plans. Under those plans, it said, Yahoo received approximately $384 million in cash proceeds in 2014, 2015 and 2016.

The order recounted the aftermath. On or about September 22, 2016, Yahoo disclosed the breach in a press release attached to a Form 8-K. The next day its market capitalization fell nearly $1.3 billion, a 3% decrease in its stock price. Verizon and Yahoo later agreed to reduce the purchase price for the operating business by $350 million, a 7.25% discount. The SEC found violations of Sections 17(a)(2) and 17(a)(3) of the Securities Act and Section 13(a) of the Exchange Act and Rules 12b-20, 13a-1, 13a-11, 13a-13 and 13a-15.

The order followed the Commission’s interpretive guidance on public company cybersecurity disclosures, dated February 21, 2018. That guidance gave an example close to the facts of the case. If a company had experienced a material denial-of-service incident, it “likely would not be sufficient” to disclose only that such an incident may occur. The company might need to discuss the incident that had happened and its consequences.

Cyber disclosure enforcement can look like a matter for securities lawyers. We read the Altaba order as an information flow problem that a technology or security leader would recognize. One fact, known to the security team within days and reported to senior managers and lawyers, was not properly assessed against the company’s public statements. The order found three kinds of statement misleading, and the company itself withdrew a fourth, its conclusion that disclosure controls worked. Each statement had a different author, a different reader and a different legal standard. All four failed for the same reason.

The order is useful beyond public companies. Publicly traded banks and bank holding companies file periodic reports with securities regulators. Health systems and public agencies that borrow in the bond markets make statements to investors in offering documents. Every regulated organization makes representations about its security in contracts, certifications and filings. The order shows how a single unrouted security finding can make many of them untrue at once.

We take the order in four parts: the three kinds of statement it found misleading and the controls conclusions the company withdrew. Each part covers a different document, written for a different purpose, so they do not overlap:

  • Risk factors: statements about what could happen, which the order found misleading because the risk had already happened.
  • Management’s discussion and analysis: statements about known trends and uncertainties, which omitted the costs the breach presented.
  • Transaction representations: statements made to a buyer in a contract, which denied breaches the company knew about.
  • Controls conclusions: statements by the principal executive and financial officers that disclosure controls worked, which the company later withdrew.

Our argument is that the failure the order described was a routing failure, and that the remedy is a map. Every organization that makes statements about its security should know which documents contain them and who must hear about a security finding before each is next signed. After the four parts, we look at why the security finding did not travel, how the finding applies across sectors and what the order did not decide. We end with the main objections and what a leader could check.

Risk factors: a risk described in the future tense after it happened

Risk factors describe what could go wrong. Yahoo’s risk factors on security breaches carried a heading that began: “If our security measures are breached, our products and services may be perceived as not being secure.” It went on to warn of users curtailing use and of significant legal and financial exposure. They said security breaches “expose us to a risk of loss of this information, litigation, remediation costs, increased costs for security measures, loss of revenue, damage to our reputation, and potential liability.”

Read alone, those sentences were unremarkable. The order’s finding turned on what the company knew when it wrote them. It said the risk factors “misleadingly suggested that a significant data breach had not yet occurred, and that therefore the company only faced the risk of data breaches.” A sentence that was accurate before December 2014 became misleading after it, without a word changing.

How the company later rewrote the risk factor

The order quoted the revised risk factor from Yahoo’s third-quarter 2016 report. It began: “Our security measures may be breached, as they were in the Security Incident and user data accessed.” The rest of the sentence kept the earlier warnings about users and legal and financial exposure. The order added italics to the clause about the incident. Yahoo also added a risk factor specific to the breach, saying the full extent of its impact and of related investigations and litigation “could be material.”

The change was small in words and large in meaning. A clause acknowledging that the risk had materialized converted a hypothetical into a fact. The February 2018 guidance had said the same thing in general terms with its denial-of-service example. In our reading, the order applied that principle to a specific set of filings.

What the February 2018 guidance asked risk factors to consider

The Commission’s February 2018 guidance listed issues that companies would find it helpful to consider in evaluating cybersecurity risk factors. The first was “the occurrence of prior cybersecurity incidents, including their severity and frequency.” Others included the probability and potential magnitude of incidents and the adequacy of preventive actions. The list also covered the aspects of the business that give rise to material risk, including third-party supplier and service provider risks.

It added the costs of maintaining protections, including insurance where applicable. It also listed reputational harm, applicable laws and regulations, and litigation, regulatory investigation and remediation costs. The guidance said companies may need to disclose previous or ongoing incidents “in order to place discussions of these risks in the appropriate context.”

The first item on that list depends entirely on the security team. Only it knows the incidents, their severity and their frequency. A risk factor process that does not ask the security team for that history each period cannot address the first consideration the Commission named.

Why risk factors are written far from the security team

Risk factors can be drafted by legal and finance staff and updated each reporting period. Their authors rely on what they are told. A security incident that has not been reported to them does not appear. That is the gap the order identified when it found that the company had no controls ensuring security reports were “properly and timely assessed to determine how and where data breaches should be disclosed.”

For a security leader, the practical point is that every risk factor about security is a statement of fact about the present as well as the future. It implies that the risk has not yet materialized in a material way. The security team is the group that knows whether that implication holds.

Risk statements outside securities filings

Organizations that never file a 10-K still describe their security risks in writing. Grant applications, responses to requests for proposals, annual reports to boards and statements to accreditors can all contain them. The same tense problem can arise. A statement that the organization “could” suffer a breach reads differently once it has. The legal consequences differ by context, and organizations should check with counsel. The drafting discipline is the same.

The MD&A requires management to discuss known trends, demands, commitments, events or uncertainties that are reasonably likely to affect liquidity or results. The order quoted the requirement from Item 303 of Regulation S-K. It found that Yahoo’s filings did not address known trends or uncertainties with regard to liquidity or net revenue presented by current or future expenses and losses related to the breach.

What the breach would have meant for the numbers

A breach of that size carries foreseeable costs: investigation, remediation, notification, legal defense and possible loss of users. The order noted that, once the breach was disclosed, Yahoo’s third-quarter 2016 report said the company expected to incur expenses including investigation, remediation and legal costs. Those expectations existed, at least in principle, from the time the theft was known. The MD&A omitted them until disclosure.

This part differs from the risk factor part in what it asks of the security team. A risk factor needs to know whether an incident happened. An MD&A needs to know what the incident is likely to cost. That requires the security team to estimate scope and consequences, and finance to translate them into expected expenses. Neither group can do it alone.

Revenue as well as expense

The MD&A requirement the order quoted covers effects on revenue as well as liquidity. Yahoo’s own risk factor warned that users and customers might curtail or stop using its services after a breach. For a business that depended on user engagement, a breach of user data was a potential revenue trend as well as a cost. The order found that the filings did not address known trends or uncertainties regarding net revenue presented by the breach.

Who estimates the cost

A cost estimate for a breach draws on several groups. The security team knows how many records and systems were affected. Legal knows the notification and litigation exposure. Finance knows how to express those as expected expenses and when to accrue them. Communications may know how customers are likely to react. An MD&A discussion requires all four to talk, which means someone must convene them.

The order’s finding suggests that the convening did not happen at the level needed. It found that management and legal staff did not properly assess the business impact of the breach. Business impact is the input the MD&A needs. Without it, the financial discussion could only describe what the company expected without the breach.

Uncertainty is not a reason to say nothing

An objection within this part is that the costs of a breach are uncertain, sometimes for years. The MD&A standard addresses uncertainty directly. It asks about known uncertainties reasonably likely to have a material effect. In our reading, uncertainty about the size of a cost is a reason to describe the range and the drivers, not a reason to leave the matter out.

Transaction representations: what the buyer was told

The third part concerns the sale of Yahoo’s operating business. The order said that during due diligence, in response to Verizon’s questions about past data breaches, Yahoo created a spreadsheet that represented it was aware of only four minor breaches exposing users’ personally identifying information. It did not disclose the 2014 theft. On a June 27, 2016 call requested by Verizon to discuss the four breaches, Yahoo again did not disclose it.

The representation in the agreement

The July 23, 2016 stock purchase agreement contained a representation about past incidents. It said that, to Yahoo’s knowledge, there had been no security breaches or unauthorized access that could reasonably be expected to have a “Business Material Adverse Effect.” The order set out the agreement’s definitions of both terms. Because the agreement was attached to a Form 8-K filed on July 25, 2016, the representation became public.

The order added a fact that sharpened the point. By June 2016, the company’s new chief information security officer had concluded that the entire user database had likely been stolen by nation-state actors and could be exposed in the immediate future. The order said this conclusion was communicated to at least one member of senior management while the sale was being negotiated.

Due diligence answers as statements

The order’s account began before the agreement, with a spreadsheet prepared in response to the buyer’s questions. It listed four minor breaches and omitted the 2014 theft. A due diligence response is drafted under time pressure, sometimes by staff far from the security team, and assembled from whatever records are at hand. In our reading, the spreadsheet shows how an incomplete record becomes an inaccurate statement when someone asks a direct question.

Organizations answer similar questions outside mergers. Customers send security questionnaires. Insurers ask about past incidents on applications. Regulators ask in examinations. Each answer is a statement about the organization’s security history. If the incident record that feeds those answers omits a known event, every answer drawn from it will too.

What the omission cost

The order recorded two measurable consequences. After disclosure, the company’s market capitalization fell nearly $1.3 billion the next day. And the purchase price for the operating business was reduced by $350 million, a 7.25% reduction. The penalty itself was $35 million. These figures come from the order and the press release; we do not attempt to separate the cost of the breach from the cost of the delay in disclosing it.

Why a contract representation counted as a public statement

A merger agreement is a private contract between parties. Its representations are written for the buyer. Once filed with the SEC as an exhibit, however, the order treated them as statements to investors. The point extends beyond mergers. Any contract an organization files publicly, or describes in a filing, can carry its security representations into the public record.

Security representations also appear in contracts that are never filed: vendor agreements, data sharing agreements, customer contracts and grant terms. They are not securities disclosures. But the same routing problem applies. A representation that no breach has occurred is only as good as the signer’s knowledge, and the signer’s knowledge depends on whether the security team’s findings reached them.

Representations given to banks and agencies

The organizations this series writes for also receive security representations. Banks receive them from service providers, health systems from business associates, and agencies from contractors. A representation that no material breach has occurred is only as reliable as the counterparty’s internal routing. A buyer of services can ask how the counterparty ensures its signers know about security findings before they sign. That question applies the Altaba order’s lesson to the buyer’s own supply chain.

Knowledge qualifiers

The representation was qualified “To the Knowledge of” the company. Knowledge qualifiers limit a representation to what specified people know. That makes the question of who knew decisive. In the order’s account, the security team and at least one member of senior management knew. In our reading, a knowledge qualifier does not protect an organization whose own security findings sit with people the signers never asked.

Controls conclusions: certifications that the system worked

The fourth part concerns the company’s statements about itself. Rule 13a-15 requires reporting companies to maintain disclosure controls and procedures. These are designed to ensure that information required to be disclosed is recorded, processed, summarized and reported on time. Each periodic report states management’s conclusion about whether those controls were effective.

Yahoo concluded in each report from 2014 through mid-2016 that its controls were effective. Its 2016 annual report reversed those conclusions for every period from December 31, 2014 through September 30, 2016. The reason it gave was deficiencies in its security incident response protocols related to the 2014 incident.

In our reading, a reversal of that kind tells investors that every certification in the period rested on a process that could not deliver the information it was meant to. In Yahoo’s case, the reversal covered every annual and quarterly period from the end of 2014 through the third quarter of 2016. The company’s explanation pointed to a single source of failure.

Why the controls failure came from incident response

The company’s own explanation located the problem in incident response, not in the disclosure process as it is drafted on paper. In our reading, that line deserves weight because it is the company’s own diagnosis. The incident response protocol determined who learned about an incident and what they did next. If the protocol stopped at the security team and selected executives, the disclosure process downstream never started.

The order made the same point in its own words. It found that Yahoo “did not maintain disclosure controls and procedures designed to ensure” that certain security reports were assessed. The reports it named were those “raising actual incidents of the theft of user data, or the significant risk of theft of user data.” In our reading, the phrase “or the significant risk” extended the expectation beyond confirmed incidents to serious warnings.

Certifications rest on what flows upward

The February 2018 guidance tied disclosure controls to the certifications that senior officers sign. It said companies should assess whether their controls ensure that relevant information about cybersecurity risks and incidents is “processed and reported to the appropriate personnel, including up the corporate ladder.” The purpose was to enable senior management “to make disclosure decisions and certifications.”

That phrase describes the route the order found missing. The officers who concluded that controls were effective were relying on a system that had not delivered the period’s central security fact to the disclosure process. Their conclusions were only as good as the route.

Trading and selective disclosure

The guidance raised two further reasons to route incident information carefully. It reminded companies and insiders of insider trading prohibitions and of the obligation to refrain from selective disclosure of material nonpublic information about cybersecurity risks or incidents. It said companies would be well served by considering how to avoid the appearance of improper trading in the period after an incident and before disclosure.

Those reminders give the routing map a second function. The people who learn of a significant incident are the people who may hold material nonpublic information. A route that tells the disclosure committee also tells the person who manages trading restrictions. The Altaba order did not address trading, and we draw no inference about it. The guidance made the connection in general terms.

What an effective control would have done

An effective control would have connected the incident response process to the disclosure committee or its equivalent. It would have defined which findings must be escalated and to whom. It would have required a documented assessment of whether the finding affected any public statement. And it would have brought in auditors and outside counsel when the finding was significant. The order noted the absence of that last step specifically.

Testing the route

A route that exists only on paper may fail when used. Incident exercises test technical response. They can also test disclosure routing, by including the general counsel’s office, finance and whoever chairs the disclosure committee. A scenario that ends with a question, such as which public statements the incident affects, tests whether the route works. Recording the answer gives the organization evidence that its controls are designed for the case the order described.

Why the security finding stalled before disclosure

Each of the four parts failed at the same junction: between the security team’s knowledge and the people responsible for a public statement. The order did not attribute motive. It described a process that did not exist.

Three reasons a finding can stall

We see three reasons such a finding can stall, drawn from the order’s facts and from the structure of organizations rather than from any claim about Yahoo’s internal debates. The first is that the security team reports to technology leadership, not to the legal or finance functions that own disclosure. Its findings travel up a different chain.

The second is that incident response focuses on containment and recovery. Its success measures are technical. Whether a public statement has become false is a different question, and an incident response plan that does not ask it will not answer it.

The third is that the people who receive the finding may not know which documents depend on it. A general counsel who hears of a breach may think of notification laws and litigation. The risk factors in next quarter’s report, a representation in a pending deal and the officers’ certification of controls are different documents in different hands.

Signals that a finding should travel

Not every security finding needs to reach the disclosure committee. A route that sends everything will be ignored. The order suggests some signals that, in our reading, should send a finding onward. One is theft of customer or user data at scale, as here. Another is attribution to a capable actor, such as the nation-state actors the order described. A third is a finding that contradicts something the organization has said publicly or in a contract.

A fourth signal is a finding that triggers legal notification duties elsewhere. If the organization must tell regulators, customers or individuals, it should also ask whether investors or counterparties have been told something different. A fifth is a serious warning without confirmation, which the order covered with its phrase about “the significant risk of theft.” Each signal is simple enough for a security team to apply without a lawyer at its side.

A single row of such a map shows how it works. The row names the document, such as the annual report’s risk factor on data security. It names the owner, such as the deputy general counsel for securities. It names the trigger, such as any confirmed theft of user data or a credible warning of one. And it names the next review date, which is the next filing.

Auditors and outside counsel as a check

The order noted that the company did not share information about the breach with its auditors or outside counsel to assess disclosure obligations. Outside advisers serve as a check on internal judgment. They see the facts from a distance and have their own professional obligations. An incident response plan that names when auditors and outside counsel are told turns their involvement from an ad hoc decision into a step.

The map as the remedy

The remedy follows from the third reason. An organization can list every document that makes a statement about its security: risk factors, MD&A, offering documents, contract representations, regulatory certifications, customer assurances. For each, it can name the owner and the trigger that should prompt a review. A security finding then has a defined route to every statement it might affect.

The map does not decide materiality. That remains a judgment for management with counsel. What the map does is ensure the judgment is made by the right people, with the facts, before the next statement is signed.

How the lesson applies across sectors

The Altaba order applied securities law to a public company. The routing problem it described applies to any organization that makes statements about its security. We look at three sectors separately, with the caveat that each organization should check with counsel on which disclosure rules apply to it.

Banks and other public financial companies

Publicly traded banks and bank holding companies file periodic reports and carry the same disclosure duties the order applied. They also make statements to supervisors in examinations and regulatory reports, and representations in bank mergers. A bank’s security finding can therefore affect its SEC filings, its supervisory communications and its deal documents at once. The map would include all three.

Bank mergers add a further document. Acquisitions among banks involve due diligence, merger agreements and, for public companies, proxy or registration statements. Each can contain statements about the target’s security history. A bank acquirer will ask the questions Verizon asked. A bank being acquired will answer them, and its answers will be compared with what it later discloses. The map for a bank would therefore include its deal documents whenever a transaction is in progress.

Supervisory communications are not securities disclosures, but the same facts feed them. A bank that tells its examiners about an incident and tells investors nothing, or the reverse, has two accounts of one event. In our reading, consistency between them is the safer course, with counsel’s advice on what each audience requires.

Health systems

Nonprofit health systems can borrow through tax-exempt bonds, issued through conduit issuers, and describe their operations in offering documents. They are not SEC reporting companies, but statements to bond investors are subject to the antifraud provisions of the securities laws, in our reading. Health systems also make security representations in payer contracts, research agreements and business associate agreements. A system that learned of a significant breach would need to ask which of those statements it affected.

For a covered entity under HIPAA, a breach of unsecured protected health information starts its own notification process. In our reading, that process is a natural trigger for the map. The same facts that require notice to patients and regulators may make a statement in a bond document, a payer contract or a grant report inaccurate. A health system that reviews those statements when it starts breach notification would close the gap the order described.

Public agencies

States, cities, counties and authorities that issue bonds also describe themselves to investors in offering documents and continuing disclosures. They make security representations in grant certifications and interagency agreements. A public agency’s incident response plan may route findings to IT leadership and the attorney’s office, but not to the finance office that prepares bond disclosures. The order’s logic suggests closing that gap.

Organizations that are privately held

A privately held company has no periodic filings, but it still signs representations in financing agreements, sale documents and customer contracts. The routing problem the order described applies to those. A private company being sold faces the same due diligence questions Verizon asked. Its answers will be tested later, as Yahoo’s were.

What the order did not decide

The order was a settled administrative proceeding. Its first footnote said the findings were made pursuant to the company’s offer of settlement and were not binding on any other person or entity. The company neither admitted nor denied them. The order is therefore a statement of the SEC’s view of these facts, not a court’s ruling.

The order found that Yahoo acted negligently in filing materially misleading reports. The charged provisions, Sections 17(a)(2) and 17(a)(3) of the Securities Act, do not require proof of intent, in our reading. The order charged no other provisions, such as Securities Act Section 17(a)(1) or Exchange Act Section 10(b), which in our reading require proof of intent. Nor did it charge any individual. The SEC’s press release said its investigation was continuing.

The order also did not set a general materiality threshold for breaches or a deadline for disclosing them. It described a case where, in the SEC’s view, no adequate assessment was made at all. Organizations reading it for a rule on timing would not find one. That question was taken up later, by rule.

Objections: materiality, timing, uncertainty and process

Four objections deserve a direct answer. Each has force, and none, in our view, undoes the argument that routing was the failure.

That not every breach is material

The first objection is that a security incident may not be material and may need no disclosure. That is correct, and the SEC’s 2018 press release reported its enforcement co-director’s statement that the agency does not second-guess good-faith exercises of judgment about cyber-incident disclosure. The order’s point was that no judgment was made at the level required. The remedy we describe routes findings to decision-makers. It does not presume the answer.

That early disclosure harms the investigation

The second objection is that disclosing too early can tip off attackers or interfere with law enforcement. Timing can matter, and later rules addressed it explicitly. But the order concerned a breach that, it said, went undisclosed in public filings for from December 2014 to September 2016, across multiple filings and a sale. Concern about timing does not explain the absence of any assessment.

That facts were uncertain

The third objection is that early in an incident, the facts are uncertain. The order’s account shows a security team that reached firm conclusions within days in December 2014 and stronger ones by June 2016. Uncertainty is real at the start of an incident. It does not last indefinitely, and the map ensures that the statements are reviewed as facts firm up.

That the map adds process

A fourth objection is that a map of security statements adds process to organizations that already have plenty. The map is a list, not a committee. Its cost is the time to compile it once and to update it when documents change. Against that cost sits the exposure the order described: three kinds of statement made misleading, and a controls conclusion later withdrawn, because one fact did not reach their authors’ assessment.

What a leader could check

The four parts translate into checks that need no new system. They are ways to see whether a security finding would reach every statement it might affect.

  • Risk factors and similar statements: whether anyone compares each statement about security risk against the incidents the security team has recorded since the last filing.
  • Financial discussion: whether finance receives an estimate of expected costs when a significant incident is confirmed.
  • Contracts: which agreements contain security representations, who signs them, and whether signers check with the security team before signing.
  • Controls: whether the incident response plan names the disclosure committee or its equivalent as a recipient of significant findings, with a documented assessment.
  • Outside advisers: when the auditors and outside counsel are told about a significant incident, and who decides.

Each check produces a record. Together, they show that the organization connected the place where security facts are discovered to the places where statements about security are made.

With hindsight

The Altaba order was followed by further enforcement on the same point. On October 22, 2024, the SEC charged four companies, Unisys, Avaya, Check Point and Mimecast, with materially misleading disclosures about cybersecurity risks and intrusions. The charges came from an investigation of companies potentially affected by the compromise of SolarWinds’ Orion software. The penalties ranged from $990,000 to $4 million. The SEC also charged Unisys with disclosure controls and procedures violations.

The SEC’s description of two of those cases was close to the Altaba finding. Its release said Unisys described its risks from cybersecurity events as hypothetical despite knowing it had experienced two SolarWinds-related intrusions involving exfiltration of gigabytes of data. It said Check Point knew of the intrusion but described cyber intrusions and risks from them in generic terms. The tense of a risk factor was again the issue.

The SEC also adopted rules that bear on that route. On July 26, 2023, it adopted rules requiring registrants to disclose material cybersecurity incidents on a new Item 1.05 of Form 8-K. The 8-K is generally due four business days after a registrant determines that an incident is material. The rules also added Item 106 of Regulation S-K, which requires annual description of processes for assessing, identifying and managing material cybersecurity risks. Item 106 also requires a description of board oversight and management’s role.

In our reading, the 2023 rule addressed the Altaba gap from two sides. Item 1.05 put a clock on disclosure once a materiality determination is made, which requires that someone make one. Item 106 required companies to describe how security risks are assessed and overseen, which makes the route from the security team to management and the board a matter of public description. The Form 8-K disclosure could be delayed if the Attorney General determined that immediate disclosure posed a substantial risk to national security or public safety.

Not every case after Altaba ended in a settled order. In October 2023, the SEC sued SolarWinds and its chief information security officer in federal court. On November 20, 2025, the SEC filed a joint stipulation to dismiss that action with prejudice. Its litigation release, citing the joint stipulation, said the decision was made “in the exercise of its discretion” and “does not necessarily reflect the Commission’s position on any other case.” We draw no conclusion about the merits from the dismissal.

The settled orders against Altaba in 2018 and the four companies in 2024 rested on a common finding, which each company neither admitted nor denied: the company knew of an intrusion while its public statements left it out or understated it. In Altaba and two of the 2024 cases, the risk factors spoke of hypothetical or generic risk. The dismissed SolarWinds case does not change what those orders found. We read it as a sign that the boundary of enforcement on cybersecurity statements remained unsettled.

For organizations outside the public company world, the developments since 2018 reinforce the same point. Public companies gained an incident disclosure clock. Contract counterparties can ask for security representations in due diligence, as Verizon did. Each of these makes the route from the security team to the person who signs a statement more consequential.

What held up from 2018 was the principle in the February guidance and the Altaba order: a risk described as possible, after it has happened, can mislead. What changed was the specificity. A security finding that might once have stalled in an inbox now has, for public companies, a rule that names the decision and a clock that starts when it is made.

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

  • SEC
  • cyber disclosure
  • disclosure controls
  • risk factors
  • data breach
  • M&A