ORDER

The memorandum disposition filed October 20,1995, is redesignated as an authored opinion by Judge Booehever.

OPINION

Butler brought an action in 1989 on behalf of the United States under the False Claims Act (“FCA”), 31 U.S.C. §§ 3729-3733 , alleging that McDonnell Douglas Helicopter Company (“MDHC”) (formerly Hughes Helicopter, Inc.) made false statements and submitted false claims to the United States Army regarding the testing and performance of communications and navigation systems on the Apache Advanced Attack Helicopter (the “Apache”). After a ten-day jury trial, the district court granted a directed verdict in favor of MDHC. Butler appeals, and we affirm.

FACTUAL BACKGROUND

Under contract from the United States Army, MDHC developed, engineered, and produced the Apache. MDHC’s work proceeded in three phases. Phase I, from 1973 to 1976, involved the development of a prototype and a “fly-off’ competition. The Army selected MDHC’s Apache as its new attack helicopter following the competition, and awarded MDHC the contract for the next phase. The Apache procurement was made through the United States Army Aviation Systems Command (“AVSCOM”), with additional technical support from other Army departments, including the United States Army Plant Representative Office (“AR-PEO”).

Phase II was a full-scale engineering development effort, which began in 1976 and ended in 1984. During those eight years MDHC developed five Apache prototypes, integrating and testing the avionics (electronic control) and navigation subsystems (“subsystems”).

In 1980, before Phase II development was complete and before the Phase II testing of the subsystems began, the Army decided for strategic and financial reasons to begin Phase III, the production of the Apache. This meant that some of the Phase II testing was curtailed or interrupted, and other Phase II tests were completed at the same time as the Phase III testing. The Phase III tests were to test an early production helicopter (the “First Article”); once that helicopter passed, the Army would formally accept delivery of Apaches produced by MDHC.

For both Phase II and Phase III testing, MDHC prepared Test Plans for the Army’s approval. The plans were reviewed by AVS-COM technical representatives, who required revisions and then recommended approval to the Army’s contracting officers, who signed the Test Plans without reading them. The Army knew that because of its decision to speed up production, the Test Plans did not include all of the testing referred to in the contract documents. Army technical representatives from ARPRO witnessed all the tests, reviewed all the Engineering Flight Test Reports, and attended pre- and post-test briefings.

After the testing pursuant to the approved Phase II Test Plan was complete, MDHC submitted a Phase II Test Report to the Army in 1983, detailing the results of the various tests on the Apache. The Army’s contracting officers approved the Phase II Test Report in early 1984, once again based entirely on the positive recommendation of the AVSCOM representatives.

Jerome F. Butler (“Butler”) worked for MDHC from 1982 to 1989. In 1983, he drafted a Phase III test plan which MDHC rejected. The Phase III Test Plan MDHC eventually prepared was approved by the Army’s contracting officers in May 1984. MDHC proceeded with testing pursuant to the Phase III Test Plan. Butler took no part in the actual testing or in the preparation of the test report. The Phase III Test Report was approved by the Army in June 1987, following AVSCOM’s recommendation.

In addition to the Phase II and III tests, before delivery, each aircraft underwent “Production Acceptance Testing” (“PAT”) by Army and MDHC pilots. The Army was always aware of the problems, or “squawks,” that arose before acceptance of each helicopter. The Army and MDHC performed yet another series of tests of the pilot’s radio system in response to field reports.

At the time of delivery and acceptance of each helicopter, the Army signed Form DD-250’s, which stated that the helicopters conformed to contract except as noted otherwise on the form or in supporting documents. From 1984 until the time of final judgment in this case in September 1993, the Army had accepted over 760 production Apaches.

Butler was convinced that his rejected draft test plan for Phase III complied with contract requirements, and the one which MDHC ultimately accepted did not. On September 29, 1989, Butler filed under seal an action claiming that MDHC had violated the False Claims Act (“FCA”), 31 U.S.C. §§ 3729-3733 , by making false statements and submitting false claims during Phase II and Phase III, related to several aspects of the testing of the avionics and navigation subsystems.

The subsystems that are the subjects of Butler’s appeal are the pilot’s radio, the copilot gunner’s radio, the automatic direction finder, and the Doppler navigation system. Generally stated, he alleged that the Test Reports falsely represented that the pilot’s and copilot gunner’s radios were demonstrated to meet specifications; that the Test Re ports falsely stated that the tests of the radios were planned in accordance with the specifications for the systems; and that the Test Reports contained false statements regarding some individual tests.

MDHC’s response was that no false statements were made, because the radios were demonstrated successfully; that the Army, through its AVSCOM and ARPRO technical representatives, approved any deviations from specification in the planned tests; and that any deviations or irregularities in the test reports were known to the Army through the AVSCOM technical representatives who were present at the tests.

The government declined to intervene. Butler proceeded with the action, which was tried to a jury. After ten days and the close of MDHC’s case, the district court granted MDHC’s motion for judgment as a matter of law, finding that there was no legally sufficient basis for a reasonable jury to find that MDHC had violated the FCA In detailed findings of fact and conclusions of law entered on August 26, 1993, the court found that given the Army’s knowledge of and access to the modifications in the testing of the subsystems, as well as the limitations of the subsystems themselves, Butler “did not present legally sufficient evidence that any allegedly false statement or claim was made with the requisite intent.” The court also found that Butler did not provide legally sufficient evidence that MDHC made any false or fraudulent statements, or that the statements made were material. Finally, the court found that the Form DD-250s introduced into evidence could not be considered statements made to the government; that many of Butler’s claims were contract disputes, not fraud claims; and that Butler failed to introduce any evidence of a connection between the allegedly false statements and any claim for payments presented to the Army.

I. Intent to knowingly make any false or fraudulent claims

We review de novo the district court’s grant of a directed verdict. Zamalloa v. Hart, 31 F.3d 911, 913 (9th Cir.1994). “ ‘A directed verdict is proper when the evidence permits only one reasonable conclusion as to the verdict.’” Id. (quoting Rudiger Charolais Ranches v. Van De Graaf Ranches, 994 F.2d 670, 672 (9th Cir.1993)). We must view the evidence in the light most favorable to Butler and draw all reasonable inferences in his favor. Zamalloa, 31 F.3d at 913 . To obtain reversal of the district court’s judgment, however, Butler must “have presented substantial evidence that would rationally support the conclusion” that MDHC violated the FCA; it is not enough if the evidence merely suggests the possibility of a violation. Browne v. McDonnell Douglas Corp., 698 F.2d 370, 371 (9th Cir.1982) (per curiam), cert. denied, 461 U.S. 930 , 103 S.Ct. 2092 , 77 L.Ed.2d 301 (1983).

A. Retroactivity of 1986 amendments to the FCA

Congress amended the FCA in October 1986, adding 31 U.S.C. § 3729 (b), which defines the mental state necessary for liability for “knowingly” presenting a false claim as actual knowledge, deliberate ignorance, or reckless disregard. It specifically provides “no proof of specific intent to defraud is required.” The statute does not indicate whether the amendment is to apply retroactively.

The district court applied the amended statute retroactively to MDHC’s statements in the Phase II and Phase III Test Reports, which were submitted in November 1983 and April 1986, before the amendments were enacted. In directing a verdict for MDHC, the court held that Butler had not produced evidence by which a jury could conclude that MDHC had acted with even the level of intent described in the 1986 amendment.

We need not decide the question of the retroactivity of § 3729(b)(1), given our holding below that the district court properly granted a directed verdict under the amended statute’s requirement that a defendant merely act with actual knowledge of falsity rather than with the specific intent to defraud. The district court’s finding that MDHC did not act knowingly would have justified a verdict in MDHC’s favor under the pre-1986 FCA

B. The Army’s knowledge of the testing problems

1. Government knowledge and defendant’s intent

The first and major basis for the district court’s grant of a directed verdict is its conclusion that:

From this finding of fact, the district court concluded: “Given the government’s knowledge of an accession at every turn to the Subsystems’ testing modifications as well as to the Subsystems’ limitations, [Butler] did not present legally sufficient evidence that any allegedly false statement or claim was made with the requisite intent.”

On appeal, Butler does not argue that this finding of fact is clearly erroneous. Instead, he argues that knowledge by the Army is not a defense to an action under the FCA alleging that a defendant “knowingly” submitted a false claim. He rests this argument on an assertion that the wrong Army personnel knew, that is, that only a contracting officer had the power to modify a government contract to allow the deviations the Army allowed in the testing. Because MDHC’s testing and reporting did not strictly comply with the specifications underlying MDHC’s contract with the Army, and because the Army technical representatives who were aware of the changes and cooperated with the revisions were not contracting officers capable of modifying the contract requirements, Butler reasons that MDHC’s “fail[ure] to adhere strictly to the contractual and military specifications” constituted a violation of the FCA.

Butler’s argument raises questions of contract interpretation rather than false claims. To the extent that he alleges that MDHC failed to comply strictly with contractual requirements in planning and reporting the testing, and that those making modifications had no authority to do so, Butler’s is a contract dispute. Such a dispute is under the jurisdiction of the Armed Services Board of Contract Appeals or the United States Court of Federal Claims. See 41 U.S.C. §§ 605 , 609.

Construing at least some of Butler’s specific allegations as FCA material, however, the issue remains whether the Army’s knowledge, through Army technical representatives, of the problems and modifications in MDHC’s testing and reporting of the Apache, defeats any inference that MDHC “knowingly” presented false claims to the government.

Before the 1986 amendments to the FCA, the answer to this question was simple: if the government already possessed the information underlying the qui tarn action, the court had no jurisdiction. 31 U.S.C. § 3730 (b)(4) (1982) provided “Unless the Government proceeds with the action, the court shall dismiss an action brought by the person on discovering the action is based on evidence or information the Government had when the action was brought.” The 1986 amendments eliminated this language, however, leaving open what would be the effect of government knowledge of the facts underlying a suit.

As government knowledge is no longer an automatic bar to suit, courts have had to decide case by case whether a FCA claim based on information in the government’s possession can succeed. In United States ex rel. Hagood v. Sonoma County Water Agency, 929 F.2d 1416 (9th Cir.1991), Hagood’s qui tarn complaint alleged that the Army had signed a water supply contract with a local agency which the agency had prepared based on inaccurate cost allocations. The complaint also alleged that the Army official signing the contract knew that accurate cost allocations had not been prepared. Id. at 1418 . The district court dismissed the com plaint as self-contradictory, finding that because the Army was aware that the cost allocations were inaccurate, and thus knew the facts that made the contract a false claim, no fraudulent intent could be shown. Id. at 1418-19 .

This court reversed, holding that the government’s knowledge of the facts underlying the false claim was not in itself an automatic defense when, as in Hagood’s case, the government’s knowledge appears merely on the face of the complaint. Id. at 1421 .

Id.

Hagood does state, as Butler urges, that the Army’s knowledge of the underlying facts is not automatically a complete defense when that knowledge appeal’s only as an allegation on the face of a complaint under the FCA. Nevertheless, Hagood left open the possibility that at the summary judgment stage or after trial, the extent and the nature of government knowledge may show that the defendant did not “knowingly” submit a false claim and so did not have the intent required by the post-1986 FCA.

This court applied Hagood in Wang ex rel. United States v. FMC Corp., 975 F.2d 1412 (9th Cir.1992). Wang alleged, among other FCA claims related to a weapons project, that the defendant’s engineering work was poor and its design was faulty. The court affirmed a summary judgment dismissing the claim, holding that because the defendant had shared with the Army all of the mistakes and limitations that later became the subject of Wang’s FCA allegations, Wang had not produced the required evidence of intent.

Id. at 1421 . Because of the dialogue with the Army, Wang’s allegations did not show the form of intent required by the Act, the “ ‘knowing presentation of what is known to be false.’ ” Id. (quoting Hagood, 929 F.2d at 1421 ). See United States ex rel. Kreindler & Kreindler v. United Technologies Corp., 985 F.2d 1148, 1157 (2d Cir.) (government knowledge may be relevant to defendant’s FCA liability), cert. denied, — U.S. -, 113 S.Ct. 2962 , 125 L.Ed.2d 663 (1993).

Under Hagood and Wang, if the district court correctly found that the only reasonable conclusion a jury could draw from the evidence was that MDHC and the Army had so completely cooperated and shared all information during the testing that MDHC did not “knowingly” submit false claims, then we must affirm the directed verdict. We examine the individual statements below, to determine whether, if they are arguably false, the government’s knowledge of their deficiencies negates MDHC’s intent.

C. Individual allegations of false claims

Butler argues that MDHC’s Phase II Test Report falsely indicated that the pilot and copilot gunner radios were successfully demonstrated and ready for production. He also argues that the Phase II Test Report falsely stated that the tests of the radios were planned in accordance with the specifications for the system and the requirements of the Test Plan. He points to what he alleges are false statements in the Phase II Test Report regarding some of the Phase II pattern testing of the pilot’s radio. Finally, he challenges the Phase III Test Report’s account of communication testing, the automatic direction finder, and the Doppler navigation system, as well as the drafting of the Phase III Test Plan.

1. Phase II Test Report

Butler claims that the Phase II Test Report, submitted to the Army in November 1983, falsely indicated that the pilot and copilot gunner radios “have been successfully demonstrated and are ready for production.” He argues that this was false because the testing of the radios was not done exactly as the contract specifications required. As MDHC points out, however, and as the district court found, this summary language about a successful demonstration does not represent that the radios met specifications, and the failure to meet specifications was detailed elsewhere in the report. We hold that the generalized statement with the added details was not the type of representation required by the statute, and so we do not address whether it was knowingly submitted.

Butler next assigns as false statements MDHC’s representations in the Phase II Test Report that the tests “were planned in accordance with the general requirements” of the relevant portions of the specification and the Phase II Test Plan, and that “[t]his report documents the results of the planned tests as modified by the program constraints described in paragraph 1.1.” Paragraph 1.1 explained that the Army and MDHC had agreed to begin Phase III before Phase II testing was complete, and that the Phase II Test Plan was not revised to reflect the planned abbreviation. Butler claims the statements are false because the radios were not successfully demonstrated, and because the Phase II Test Report does not provide the results of the tests in the Test Plan. Yet these statements say nothing about successful demonstration, and actually disclose what Butler claims they conceal: that the Phase II Test Report will not be in exact accord with the Phase II Test Plan. We hold that these statements are not false, and again decline to decide whether they were knowingly submitted.

2. Phase II Pattern Tests

Butler next identifies as false statements the Phase II Test Report’s general representation that pattern tests of the pilot’s radio were done at 10 to 12 nautical miles and 1200 feet altitude. Butler argues that at least two test flights were actually done at closer distances. Yet he also alleges that the Army witnesses’ reports stated that the tests were flown at closer distances, and MDHC points to testimony that the tests were the subject of discussion between MDHC and the Army, and the Army’s AVS-COM technical representatives knew of and approved the short tests.

We hold that the Army knew that the summary statement of the distances at which the pattern tests were conducted was not strictly accurate as to all the tests, and that this discrepancy was the subject of dialogue between the Army and MDHC. The only reasonable conclusion is that this was not a “knowingly” false statement, as the noncomplying tests were known to and approved by the Army.

Butler also points to discrepancies within the report itself, such as the substitution of Flight #827 for #822. He labels these as “falsified results from tests of the FM band for the pilot’s radio”, but does not demonstrate why they are knowingly false statements rather than clerical errors. He also labels as a false claim the pilot’s radio patterns in the Phase II Test Report, because there were more data points in the Test Report than had appeared in an interim report a year earlier. Butler argues that the discrepancy must mean that data was falsified, but as MDHC points out, the Army had both the interim and final report, and the data are comparable in shape and readings despite the differences in the number of data points. We hold that Butler did not present evidence that rationally supports the conclusion that MDHC falsified data. Even if the discrepancy constituted a false statement, the Army knew of the discrepancy, and so MDHC did not “knowingly” submit a false claim.

Butler then complains that given the ambient noise at the ground station, the patterns for Flight # 370, attached to the Phase II Test Report, must be false. He cites no evidence to support this, and MDHC points to testimony that the recorded points were all above the ambient noise level. Butler next questions the accuracy of a table charting a range test result in the Phase II Test Report, but as MDHC points out, the Army had the flight test reports which noted a garbled signal, and the table itself stated that several frequencies were less than satisfactory. We hold that these statements are not false.

3.Phase III Tests

Butler claims that MDHC “rigged” a test in the Phase III Test Plan, which describes a one-way test for the radios rather than the two-way test required by the specification. MDHC did perform two-way tests, however, so there was no failure to perform a required test. Butler’s statement that an Army flight test witness was “manipulated” by an MDHC employee into reporting favorable test results is not strictly consistent with the witness’s testimony that he “took [the MDHC employee’s] word for it” and is in any event irrelevant, as the Phase III Test Report disclosed mixed results for the test.

Butler also turns to the copilot gunner’s radio, claiming that the Phase III Test Report falsely stated that it passed all tests and was ready for production. He apparently objects to the way in which the tests were documented, but this does not indicate that the test reports were in any way false.

Butler then claims that the Phase III Test Report contained a “rigged” test for the Automatic Direction Finder (“ADF”). His argument that the ADF “could not possibly” have complied does not identify an alleged false claim; his complaint about the Phase III Test Plan’s criteria is a rehash of his general complaint about the Test Plan; and his description of the problems that the Army found in its later production testing provides no evidence of falsity in the Phase III Test Report.

Butler’s arguments regarding the Doppler Navigation System similarly fail to provide any evidence on which a jury could find a false claim. He identifies as false MDHC’s statement that the Doppler did not have to be tested during the Phase II navigation subsystem tests. The Doppler, however, was excluded from the Phase II tests as the Test Plans reflect. Differences between MDHC and the Army about how the Doppler was to be tested were openly aired and eventually resolved. Butler complains that the Doppler never passed the Phase III tests, but overlooks that the very test results he cites indicate that the Doppler failed.

4.Drafting of the Phase III Test Plan

Butler argues that the Phase III Test Plan, which was selected over Butler’s rejected version, was improperly structured and did not properly provide for data collection. This is not a FCA claim, but a contract claim. The improper interpretation or unauthorized amendment of a contract, without more, does not constitute a false claim for payment.

5.Conclusion

We affirm the district court’s directed verdict on the ground that the only reasonable conclusion a jury could draw from the evidence was that all the false statements alleged by Butler as the basis for his allegations of false claims either were not false, or were not made with the requisite knowledge under the FCA.

II. Other issues

A. Materiality

As an alternative ground for its directed verdict, the district court found that there was no legally sufficient basis for a reasonable jury to find that if MDHC did make any false statements, those false statements were material. Butler argues that the FCA does not require materiality. Because we affirm the directed verdict on the basis that the only reasonable conclusion a jury could draw from the evidence was that the false statements Butler identified were not false or were not made with the required knowledge, we need not reach this issue.

B. Form DD-250S

The district court found that there was no legally sufficient basis to find that the Form DD-250s introduced by Butler constituted claims under the FCA, because “A signed DD-250 is a representation by a government official that a product conforms to contract specifications.” Butler urges that the DD-250s constitute representations by MDHC and thus were “claims” under the FCA because MDHC prepared them, and because they were prerequisites to submitting an invoice.

A DD-250 is a form supplied by the government which the contractor fills in, identifying the contract by number and providing a description of the item; the government then signs the form. The form has a variety of uses, and can be used as an invoice. The DD-250s in this ease were not submitted as invoices (and, in fact, MDHC submitted separate invoices). Instead, they were used as receiving forms, “completed by a government official ... indicat[ing] that the goods had been received and had been inspected by a government official.” United States v. Wertheimer, 434 F.2d 1004, 1005 (2d Cir.1970). Nor did Butler introduce into evidence any proof of misrepresentation in the supporting documents that may have accompanied the DD-250s.

The United States has submitted an ami-cus brief urging this court to affirm the district court’s decision regarding the DD-250s on the narrow ground that the DD-250s in this case were not invoices, and the district court found no other evidence of failure to meet contract specifications.

In United States v. Cannon, 41 F.3d 1462 (11th Cir.1995), cert. denied, — U.S. -, 116 S.Ct. 86 , 133 L.Ed.2d 44 (1995), the Eleventh Circuit reversed a conviction under 18 U.S.C. § 1001 on the ground that a contractor did not lie to the government by submitting a Form DD-250, even though the parts supplied did not conform to specifications. “Nowhere on the DD-250 does [the defendant] or anyone else at [the company] certify that the parts supplied conform to the contract. Rather, the government [quality assurance representative] signs the form signifying acceptance and conformance of the goods.” Id. at 1468 . The court rejected the government’s argument that the defendant caused the government representative to make a false statement by “presenting the DD-250 forms, representing to [the government] that the contracts had been performed to specifications, and having the [government representative] sign the DD-250, accepting the items and certifying that they met the contract specifications.” Id. Without more, i.e. a false supporting document, no reasonable trier of fact could find that the defendant lied to the government or submitted a false document: “Although the [material] did not conform, the documents did not lie. It was through the failure of the [government representative] to perform an adequate review that the nonconforming material was certified.” Id. at 1468-69 .

Cannon’s reasoning is directly applicable to the facts in this case. The DD-250 was not a representation by MDHC, because the government, not MDHC, certified on the form that the goods conformed to contract.

We affirm the district court’s ruling regarding the DD-250s. Under these circumstances, where the DD-250s were not used as invoices, and the qui tam plaintiff introduced into evidence no false supporting documents, the DD-250s were not “claims” under the FCA.

C. MDHC’s motion to dismiss the complaint

In its answering brief, appellee MDHC argues that the case should have been dismissed for lack of jurisdiction because Butler violated the FCA’s sealing requirements when his attorney appeared on nationwide television and discussed the complaint’s allegations. MDHC moved to dismiss the complaint on this ground in the district court.

MDHC argues that the in camera and sealing requirements are jurisdictional, relying on Erickson ex rel. United States v. American Inst. of Biological Sciences, 716 F.Supp. 908 (E.D.Va.1989), and that Butler’s attorney’s television appearance violated the requirements. This court recently has held that the sealing requirements of 31 U.S.C. § 3730 (b)(2) are not jurisdictional. United States ex rel. Lujan v. Hughes Aircraft Co., 67 F.3d 242, 246 (9th Cir.1995). Because we have affirmed the district court’s ruling on its merits, we need not decide whether the disclosure violated the in camera requirements.

AFFIRMED.