with whom MEAD, J., joins, dissenting.
[¶ 22] I respectfully dissent because I believe that the functional equivalency analysis requires us to conclude that the records of the investigatory panel organized by the Attorney General are subject to the Freedom of Access Act, 1 M.R.S. §§ 401-412 (2007).
*986A. Functional Equivalency Analysis
[¶ 23] To determine whether a private entity qualifies as a public agency under the Freedom of Access Act, we employ the four-part functional equivalency test developed by the Connecticut Supreme Court. See Town of Burlington v. Hosp. Admin. Dist. No.1, 2001 ME 59, ¶ 16, 769 A.2d 857, 862-63 (citing Conn. Humane Soc’y v. Freedom of Info. Comm’n, 218 Conn. 757, 591 A.2d 395, 397 (Conn.1991)). This test requires us to consider: (1) whether the entity performs a governmental function; (2) whether the government funds the entity; (3) the extent of governmental involvement or control; and (4) whether the government created the entity. Id. ¶ 16, 591 A.2d 395, 769 A.2d at 863; Conn. Humane Soc’y, 591 A.2d at 397; see also Dow v. Caribou Chamber of Commerce & Indus., 2005 ME 113, ¶ 12, 884 A.2d 667, 670 (quoting verbatim the test outlined in Town of Burlington). An entity need not strictly conform to each factor. Instead, “[a]ll relevant factors are to be considered cumulatively, with no single factor being essential or conclusive.” Ry. Labor Executives’ Ass’n v. Consol. Rail Corp., 580 F.Supp. 777, 778 (D.D.C.1984).
1. Performing a Governmental Function
[¶ 24] The investigatory panel organized by the Attorney General was clearly performing a traditional government function-the internal investigation of allegations of prosecutorial and law enforcement misconduct. The fact that an investigation is conducted through an outside source does not necessarily change the public nature of the investigation. See Cyr v. Madawaska Sch. Dep’t, 2007 ME 28, ¶¶ 3-4, 12, 916 A.2d 967, 969, 971 (holding that an outside investigation of a school matter did not make portions of the resulting report any less a confidential employee evaluation); Lewiston Daily Sun, Inc. v. City of Auburn, 544 A.2d 335, 335-36 (Me.1988) (finding an independent, volunteer investigatory committee subject to the Act). In reaching the opposite conclusion, the Court minimizes several pertinent facts.
[¶25] Before creating the panel, the Attorney General’s office discussed whether there was any effective way to conduct an independent review within the state government of the alleged misconduct in the Dechaine case and eventually concluded there was not. The Attorney General asked a jurist and two private attorneys, rather than his own staff of attorneys and investigators, to conduct this investigation “in order to ensure continued public confidence” in the state’s law enforcement agencies. To facilitate this internal investigation, the Attorney General not only provided the panel access to his entire file in the matter, portions of which are explicitly designated confidential by the Legislature,1 but also ordered the personnel involved in the Dechaine prosecution to cooperate in the panel’s investigation. Given the panel’s unfettered access to interview, at its discretion, the prosecutors and investigators involved in the prosecution of a high profile murder case, the panel was afforded a truly unique window through which to view and assess the inner workings of the Attorney General’s Department, a view that is otherwise not available to private interest groups or the public.
[¶ 26] Because an investigation initiated by the Attorney General of the internal workings of his staff in connection with a specific prosecution is a core function of *987his office, the first criterion of the functional equivalency analysis weighs in favor of finding that the panel was performing a governmental function. The Court reaches the opposite conclusion, focusing its analysis of the first criterion on the manner by which the investigatory panel was created. I disagree with this approach for two reasons.
[¶27] First, the government’s role in the creation of the entity is a separate criterion — the fourth — of the functional equivalency test. The Court’s approach conflates the first and fourth criteria, resulting in the same facts being considered twice.
[¶ 28] Second, the Court’s analysis improperly focuses on the limitations on the statutory term “public proceedings,” 1 M.R.S. § 402(2)(F), as part of its analysis in a case involving “public records,” 1 M.R.S. § 402(8). We have long recognized that the Legislature has decoupled the term “public proceedings” from “public records.” See Moffett v. City of Portland, 400 A.2d 340, 341 n. 4 (Me.1979) (describing the statutory evolution of the term “public records” and its expansion beyond merely the records and minutes of public proceedings). Furthermore, “[t]he Legislature has declared that the [Freedom of Access Act] ‘shall be liberally construed and applied to promote its underlying purposes and policies’ ” and we therefore strictly construe all statutory exceptions to public disclosure. Cyr, 2007 ME 28, ¶8, 916 A.2d at 970 (quoting 1 M.R.S. § 401). The limitations regarding advisory groups contained in the statutory definition of “public proceedings,” are not repeated in the list of statutory exemptions to the definition of “public records.” See 1 M.R.S. § 402(2), (3). Accordingly, these limitations do not control our interpretation of “public records” and the Court’s use of these exemptions to narrow the scope of “public records” runs contrary to both the statutory text and legislative intent. See 1 M.R.S. § 401.2
2. Government Funding
[¶ 29] The manner in which the investigation was financed, on the other hand, weighs against a finding that the panel qualifies as a government agency under the Act. The panel members received no compensation for their activities, and the law firm of one of the members funded the panel’s administrative costs. The Attorney General’s Department otherwise provided only incidental, logistical support in arranging interviews of those persons the panel wished to interview.
3. The Extent of Government Involvement or Control
[¶ 30] The extent of government involvement or control criterion weighs in favor of finding the panel acted as a public agency. Although the Attorney General provided only incidental, logistical support and, once formed, the panel acted independently of his control, the Attorney General delineated the scope of the panel’s work. The Court’s analysis, in my view, minimizes key facts regarding this criterion.
[¶ 31] In his October 23, 2004, letter formalizing the panel members’ appointments, the Attorney General identified five specific allegations and instructed them that, “[y]ou are asked to investigate these [five specific] allegations only” and instructed them to report their findings di*988rectly to him.3 After completing its review, the panel did not release the report to the general public. Instead, it followed the Attorney General’s instructions and provided the report to him directly. The Attorney General later released the report himself.
[¶ 32] The panel acknowledged the Attorney General’s control over the scope of its investigation in its letter detailing its findings:
Our sole purpose was to investigate the allegations detailed above, and to advise you, after our independent review, whether we found any of the allegations made against your office or law enforcement officers had any substantive merit,
[f 33] Prior to adopting our present functional equivalency analysis, we addressed the application of the Freedom of Access Act to a volunteer investigatory committee created by a city council and mayor in Lewiston Daily Sun, Inc. v. City of Auburn, 544 A.2d 335 (Me.1988). In concluding that the Act applied to the committee, we focused in part on the substantial links between the committee and the mayor. Id. at 338. Like the Attorney General, the mayor created the committee, charged it to investigate a particular issue, and asked the committee to provide him with a recommendation. Id. Although the mayor recommended that the committee interview certain individuals, he told the committee that it could interview whomever it pleased. Id.
[¶ 34] As with the mayor in Lewiston Daily Sun, the Attorney General exercised no direct control over the panel during the course of its investigation. He did, however, exercise substantial control over the panel’s formation, its composition, its access to departmental resources, the scope of its inquiry, and the distribution of its findings. Therefore, this factor weighs in favor of a positive finding that the panel acted as a public agency.
4. Created by The Government
[¶ 35] Finally, the fourth criterion— whether the government created the entity — weighs in favor of finding that the panel acted as a public agency. Although we previously described this criterion as an inquiry into whether the entity was “created by private or legislative action,” see Town of Burlington, 2001 ME 59, ¶ 16, 769 A.2d at 863, the functional equivalency test is not so narrow. The decisions from other jurisdictions on which we have relied in describing the functional equivalency test establish that the fourth criterion is “whether the entity was created by the government,” not simply whether the entity owes its existence to legislative action. See Conn. Humane Soc’y, 591 A.2d at 397, cited in Town of Burlington, 2001 ME 59, ¶ 16, 769 A.2d at 863; Telford v. Thurston County Bd. of Comm’rs, 95 Wash.App. 149, 974 P.2d 886, 893 (1999).
*989[¶ 36] This broader understanding of the fourth criterion is necessary because the Freedom of Access Act is not restricted to entities created by legislation. See 1 M.R.S. § 402(2)(F) (defining “public proceedings” subject to the Freedom of Access Act as including “[a]ny advisory organization ... established ... by Executive Order issued by the Governor”); 1 M.R.S. §§ 402(3), 412(4) (establishing that the Act applies to various public officials, including constitutional officers not created by legislative action). Although the Court describes its analysis as an examination of whether the entity at issue was created by “legislative action,” it proceeds to expand the scope of its inquiry into whether there has been an “Executive Order of the Governor” or other “official action.” See supra ¶ 1.
[¶ 37] In this instance, although there was no legislative mandate, the panel was clearly created by official governmental action. The Attorney General, a constitutional officer acting in his official capacity, conceived of the idea of an independent investigatory panel, appointed its members, delineated the scope of its work, and provided it unfettered access to his staff.
5. Balancing of the Four Factors
[¶ 38] Of the four factors, three support the conclusion that the panel is subject to the Freedom of Access Act. The second factor, which requires us to consider the manner in which the panel was funded, weighs against a positive finding. Nevertheless, the voluntary nature of the panel and the absence of governmental financial support is not dispositive of the matter because although all four factors must be “considered and weighed, an entity need not strictly conform to each of the factors.” Dow, 2005 ME 113, ¶12, 884 A.2d at 670 (quotation marks omitted). Numerous boards, commissions, and committees of this State receive no financial compensation for their public service but may be subject to the Freedom of Access Act. See 5 M.R.S. § 12004-1 (2007) (listing various statutorily created boards for which compensation is not authorized).
[¶ 39] I am ultimately persuaded by the circumstances presented to give the greatest weight to the first factor — whether the entity performed a governmental function — because of the unique nature of the governmental function the panel performed. See Portland Water Dist. v. Town of Standish, 2006 ME 104, ¶23, 905 A.2d 829, 835-36 (finding a water district to be a governmental entity due in part to its “uniquely governmental functions”). The power to prosecute rests exclusively with the State and is a core function of sovereign authority. See United States v. Lara, 541 U.S. 193, 197-98, 124 S.Ct. 1628, 158 L.Ed.2d 420 (2004). Any investigation into the exercise of that power initiated by the Attorney General, Maine’s chief law enforcement officer, is necessarily a derivative expression of this unique authority. In this case, the panel’s investigation sprang directly from the Attorney General’s exclusive responsibility for the “direction and control of all investigations and prosecution of homicides,” 5 M.R.S. § 200-A (2007). The fact that three volunteers appointed by the Attorney General conducted the investigation, and not paid, professional members of his department, offers no rational basis to diminish the public’s right, pursuant to the Freedom of Access Act, to seek access to the materials resulting from the investigation.
B. Conclusion
[¶ 40] Measured against our previously established criteria, the undisputed facts lead to the conclusion that the panel acted as an arm of the Department of the Attorney General. All materials generated in *990the course of the panel’s investigation, not otherwise covered by an exception, see, e.g., 5 M.R.S. § 7070(2) (2007); 16 M.R.S. § 614 (2007); 30-A M.R.S. § 2702(1)(B)(5) (2007), should be subject to the Freedom of Access Act. The panel and its members therefore should be obligated to respond to Moore’s request pursuant to 1 M.R.S. § 409(1).
. The Legislature, in creating an exception for this case to the pre-1995 exception to 16 M.R.S. § 614 (2007), explicitly left certain aspects of the file confidential and not subject to release under section 614. See P. & S.L. 2003, ch. 18.
. As we stated in Town of Burlington v. Hospital Administrative District No. 1, 2001 ME 59, ¶ 14 n. 7, 769 A.2d 857, 861: "Although the list of these entities is not directly applicable to this case because this case concerns records, not proceedings, the list is illustrative of the breadth of organizations covered by [the Freedom of Access Act].”
. The five allegations are as follows:
• Following their initial investigation, law enforcement officers altered their notes and/or reports to falsely attribute incriminating statements to Dennis Dechaine.
• Prosecutors misled the jury with respect to [the victim’s] time of death.
• At the time of trial, prosecutors and law enforcement officers had information about an alternative suspect which they should have shared, but did not share, with defense counsel.
• In 1992, law enforcement officers, with the approval of prosecutors, inappropriately destroyed physical evidence including a rape kit as well as hairs and fibers discovered at the scene where [the victim] was found.
• Prosecutors inappropriately failed to notify the court and defense counsel of a consultant’s opinion regarding the reliability of an outside laboratory and DNA tests conducted in 1993.