concurring in part and dissenting in part.
I agree with the majority that cases involving qualified immunity and a § 1983 action predicated on the state-created-danger doctrine require that we answer three questions. We must (1) identify the constitutional right and determine whether it was violated, Saucier v. Katz, 533 U.S. 194, 200, 121 S.Ct. 2151, 150 L.Ed.2d 272 (2001); (2) determine whether that right was clearly established at the time the violation occurred, id.; and (3) determine whether the plaintiff established state liability for a state-created danger, Kallstrom v. City of Columbus, 136 F.3d 1055, 1066 (6th Cir.1998). Because I believe the majority reads Kallstrom too narrowly, thereby answering the first question wrongly and ignoring the remaining questions, I respectfully dissent from Part III of the majority’s opinion.
I.
A. Constitutional Violation
The majority contends that the constitutional right at issue here is somehow different from the constitutional right at issue in Kallstrom. Kallstrom involved the release of three undercover police officers’ personal information — including their names, addresses, phone numbers, social security numbers, and family members’ names — to the attorney of several members of the violent “Short North Posse.” Id. at 1059. The Kallstrom Court held that the City’s improper release of such sensitive personal information rose to constitutional dimensions, implicating the substantive due process component of the Fourteenth Amendment, because the disclosure of the officers’ personal information “created a very real threat to the officers’ and their family members’ personal security and bodily integrity, and possibly their lives.” Id. at 1063-64. I am not persuaded by the majority’s conclusion that the right at issue in this case is in any way different than the right at issue in Kallstrom.
Here, the Defendants improperly disclosed several corrections officers’ names, birth dates, and social security numbers to an inmate population that otherwise would not have had access to such confidential information. Armed with this information, inmates wreaked havoc on the officers’ lives by using the officers’ social security numbers to obtain other confidential information, including each officer’s home address and family members’ names. Specifically, inmates have made numerous death threats against the officers and their families, referencing the officers’ home addresses, social security numbers, and family members by name. In a substantial step towards making good on these threats, inmates have sent confederates on the outside to the officers’ homes. Evidence of this includes prison officials intercepting incoming prison mail containing photos of Melissa Barber’s house and car, and, most frighteningly, inmates accurately describing the officers’ children in taunts. Because I believe it is clear that the disclosure of the corrections officers’ personal information “created a very real threat to the officers’ and their family members’ personal security and bodily integrity, and possibly their lives,” id. at 1063-64, it is difficult to see how this case does not involve a violation of the same *462constitutional right that we described in Kallstrom. Indeed, Judge Cook’s concurrence implicitly suggests as much. See Cook Op. at 458 (“[I]t is worth rethinking Kallstrom.”).
The majority unhappily acknowledges that Kallstrom created a privacy right protecting an individual’s sensitive personal information — -one it describes as “narrowly tailored” and “limited to circumstances where [1] the information disclosed was particularly sensitive and [2] the persons to whom it was disclosed were particularly dangerous vis-a-vis the plaintiffs.” Maj. Op. at 456. The majority is careful to point out, however, that this right “is virtually unique among courts of appeals.” Id. at 455. Of course, what our sister circuits have done is irrelevant as this Circuit has clearly recognized the right; indeed, we are bound by the prior panel’s opinion in Kallstrom. See, e.g., Dingle v. Bioport Corp., 388 F.3d 209, 215 (6th Cir.2004) (“A panel of this Court cannot overrule the decision of another panel.”). In attempting to distinguish Kallstrom, the majority explains that social security numbers are not sensitive enough, nor were the inmates dangerous enough. I disagree on both points.
1. Personal Information
The majority concludes that social security numbers are not “tantamount to the sensitive information disclosed in Kallst-rom.” Maj. Op. at 456. The majority highlights the fact that in Kallstrom “[t]he district court did not make any explicit findings with respect to whether disclosure of ... social security numbers ... put the officers at substantial risk of serious bodily harm[,]” and therefore a remand was necessary for the district court to “consider the extent to which the release of this information jeopardized the officers’ personal security, and whether the threat, if any, implicated the officers’ constitutionally protected interests in privacy and bodily integrity.” Id. at 455 (quoting Kallstrom, 136 F.3d at 1063 n. 2). Inexplicably, however, the majority holds that social security numbers today are not sufficiently sensitive as a matter of law. Adherence to Kallstrom would seem to require a remand on this point.
In any event, in the nearly ten years since Kallstrom was decided, with the growth of the internet and ubiquitous online databases, social security numbers have become, if anything, more sensitive. See generally, e.g., Lynn M. LoPucki, Human Identification Theory and the Identity Theft Problem, 80 Tex. L.Rev. 89 (2001). Indeed, armed with a social security number and an internet connection, anyone can obtain an individual’s credit report, which, at a minimum, contains the individual’s name, address, phone number, birth date, employer, and spouse’s name, in addition to credit information and public-record information. See, e.g., Federal Trade Commission: Building a Better Credit Report, http://www.ftc.goVbep/edu/pubs/consumer/ credit/cre03.shtm (last visited July 30, 2007). This is materially indistinguishable from the information we deemed to be sufficiently sensitive in Kallstrom. See Kallstrom, 136 F.3d at 1069-70 (“We hold that because disclosure of the officers’ addresses, phone numbers, and driver’s licenses, as well as the names, addresses, and phone numbers of their family members, placed the officers and their families at substantial risk of serious bodily harm, the prior release of this information encroached upon their fundamental rights to privacy and personal security under the Due Process Clause of the Fourteenth Amendment.”).
Therefore, given that today a social security number is a veritable key to an individual’s most sensitive personal infor*463mation, I am unpersuaded by the majority’s analysis that social security numbers are less sensitive than the information disclosed in Kallstrom, and the majority’s implicit conclusion that, in the almost ten years since we decided Kallstrom, social security numbers have become less sensitive such that a remand on this point is no longer warranted.
2. Relationship Between Inmates and Corrections Officers
The majority concludes that the threat of retaliation from the inmates was not apparent. That is, the relationship here between the inmates and the corrections officers “is not defined by the clear animosity apparent in Kallstrom.” Maj. Op. at 457. Ignoring the benefit of hindsight (which confirms that the informationally empowered inmates did in fact retaliate against the officers and their families, substantially putting them at risk), the relationship here between the corrections officers and inmates, particularly those housed at a Level VI “Super Max” facility, is certainly characterized by animosity. Indeed, in Michigan; Level VI facilities are reserved only for prisoners with substantial behavioral problems who cannot be safely housed at other facilities. Cain v. Mich. Dept. of Corr., 451 Mich. 470, 548 N.W.2d 210, 212 n. 4 (1996); see also, e.g., Anthony v. Gilman, No. 4:03-CV-87, 2006 WL 222842, at * 1 (W.D.Mich. Jan.26, 2006) (“IMAX [is] known as a ‘Level 6’ facility, or ‘Supermax,’ meaning that it house[s] the worst of the worst offenders in terms of security.”).
I cannot conclude that the threat of retaliation was any less severe here than in Kallstrom. Clashes between inmates and guards are nothing new. See, e.g., Morgan v. Ward, 699 F.Supp. 1025, 1055 (N.D.N.Y.1988) (“The realities of a maximum facility correctional institution are far removed from the peace of a judge’s chambers, and the court is fully aware that the exchange of verbal insults between inmates and guards is a constant, daily ritual observed in this nation’s prisons.” (citation and quotation marks omitted)). Corrections officers must be ever vigilant of constant, and often innovative, threats to their safety, ranging from verbal assaults to excrement bombs to stabbings. See, e.g., Rust v. Grammer, 858 F.2d 411, 412 (8th Cir.1988) (“For several months, inmates had been setting fires and throwing food, urine, and feces into the gallery and onto the guards.”); Bruscino v. Carlson, 854 F.2d 162, 165 (7th Cir.1988) (“Inmates have attacked other inmates and guards with a homemade bomb, with a light bulb, with a padlock, with a sharpened pencil wielded as a knife, with a sharpened toothbrush, with feces, with a chair, with a mop wringer, with a home-made mallet, and with a bucket of boiling water, as well as with the usual zip guns and shanks.”). It does not take much to imagine what inmates housed at a maximum-security facility would do when handed the social security numbers of the corrections officers who guard them. They would do exactly what they did here — “create[ ] a very real threat to the officers’ and their family members’ personal security and bodily integrity, and possibly their lives.” Kallstrom, 136 F.3d at 1063. Indeed, the Defendants concede that social security numbers are exactly the sort of information that should never have been disclosed to the inmates. I cannot join the majority’s holding that the adversarial relationship here between the corrections officers and the inmates was not sufficiently analogous to the animosity apparent in Kallstrom.
Because the facts of this case are substantially identical to Kallstrom, and because I believe the majority has unpersua-*464sively distinguished Kallstrom, I would hold that the improper release of the corrections officers’ sensitive personal information “encroached upon their fundamental rights to privacy and personal security under the Due Process Clause of the Fourteenth Amendment,” where its disclosure “placed the officers and their families at substantial risk of serious bodily harm.” Id. at 1069-70. Moreover, the Defendants have not put forward, and I cannot fathom, a legitimate state interest in disclosing this personal information to the inmates. See id. at 1070; J.P. v. DeSanti, 653 F.2d 1080, 1091 (1981). Accordingly, I now answer the remaining two questions, which the majority found unnecessary to address.
B. Clearly Established Right
I have little trouble concluding that the constitutional right here was clearly established. “In order to conclude that the right [that] the official allegedly violated is ‘clearly established,’ the contours of the right must be sufficiently clear that a reasonable official would understand that what he is doing violates that right.” Anderson v. Creighton, 483 U.S. 635, 640, 107 S.Ct. 3034, 97 L.Ed.2d 523 (1987). “[Qualified immunity is intended to provide government officials with the ability ‘reasonably [to] anticipate when their conduct may give rise to liability for damages.’ ” Id. at 646, 107 S.Ct. 3034 (quoting Davis v. Scherer, 468 U.S. 183, 195, 104 S.Ct. 3012, 82 L.Ed.2d 139 (1984)). Given the striking similarity between the facts of this case and the facts in Kallstrom, “in light of pre-existing law the unlawfulness [would have been] apparent.” Id. Indeed, we held the very action in question here unlawful in Kallstrom. That is, a reasonable state official would have been aware that releasing corrections officers’ personal information — i.e., names, birth dates, and social security numbers — to inmates at a Level VI, “Super Max” facility could result in the deprivation of the officers’ due process rights to personal security and bodily integrity. Kallstrom, 136 F.3d at 1067; cf. Bloch v. Ribar, 156 F.3d 673, 687 (6th Cir.1998) (“In light of our ruling in the present case, however, public officials in this circuit will now be on notice that such a privacy right exists. Therefore, any future violation will not allow an official ... to claim the lack of reasonable notice that is necessary to sustain a defense of qualified immunity.”).
C. State-Created Danger
To bring a state-created-danger claim, the individual must show: “(1) an affirmative act by the state [that] either created or increased the risk that the plaintiff would be exposed to an act of violence by a third party; (2) a special danger to the plaintiff wherein the state’s actions placed the plaintiff specifically at risk, as distinguished from a risk that affects the public at large; and (3) the state knew or should have known that its actions specifically endangered the plaintiff.” Cartwright v. City of Marine City, 336 F.3d 487, 493 (6th Cir.2003) (citing Kallstrom, 136 F.3d at 1066). As we have stated repeatedly, “state officials may violate the Due Process Clause when their affirmative actions directly increase the vulnerability of citizens to danger or otherwise place citizens in harm’s way.” Ewolski v. City of Brunswick, 287 F.3d 492, 509 (6th Cir.2002) (citing DeShaney v. Winnebago County Dep’t of Soc. Servs., 489 U.S. 189, 201, 109 S.Ct. 998, 103 L.Ed.2d 249 (1989)). In other words, the issue is whether the state officials did anything “affirmative” to “embolden” the person causing harm to another. Jones v. Reynolds, 438 F.3d 685, 703 (6th Cir.2006).
1. Afirmative Act
“Liability under the state-created-danger theory is predicated upon affirmative *465acts by the state [that] either create or increase the risk that an individual will be exposed to private acts of violence.” Kallstrom, 136 F.3d at 1066. Because it is often difficult to distinguish action from inaction, “[rjather than focusing on the often metaphysical question of whether officer behavior amounts to affirmative conduct or not, we have focused on ‘whether [the victim] was safer before the state action than he was after it.’ ” Koulta v. Merciez, 477 F.3d 442, 445-46 (6th Cir.2007) (quoting Cartwright, 336 F.3d at 493).
There is little doubt that Lowery’s release of the corrections officers’ personal information substantially increased the officers’ and their families’ vulnerability to private acts of vengeance. As in Kallst-rom, “if [Lowery] had not acted, then the attackers would not have access to this information and attacks would not be facilitated.” McQueen v. Beecher Cmty. Schs., 433 F.3d 460, 466 (6th Cir.2006) (citing Kallstrom, 136 F.3d at 1067).
On the other hand, Sibert’s role in the release of the corrections officers’ personal information was limited to preparing the Internal Affairs investigative report that contained the corrections officers’ personal information, which Lowery ultimately released. Sibert simply investigated the prisoners’ charges and compiled a report for internal consumption. He played no role in determining what information would be released to the prisoners, or whether the report should be released at all. The corrections officers were no less safe after Sibert acted than before. It was Lowery’s imprudent release of this report that created the danger.
Because Sibert did not act affirmatively to create or increase the risk that the corrections officers’ would be exposed to private acts of violence, summary judgment in favor of Sibert was appropriate. The analysis therefore proceeds addressing only Lowery.
2. Special Danger
To establish a “special danger,” the Plaintiffs must show that “the state’s actions placed the victim specifically at risk, as distinguished from a risk that affects the public at large.” Jones, 438 F.3d at 690 (citing Kallstrom, 136 F.3d at 1066). The Plaintiffs have no difficulty demonstrating a special danger. As in Kallst-rom, Lowery’s release of the corrections officers’ personal information “placed the personal safety of the officers and their family members, as distinguished from the public at large, in serious jeopardy.” Kallstrom, 136 F.3d at 1067.
3. State Culpability
Finally, the Plaintiffs “must demonstrate that the state acted with the requisite culpability to establish a substantive due process violation under the Fourteenth Amendment.” Ewolski, 287 F.3d at 510. The state’s conduct must be “so ‘egregious’ that it can be said to be ‘arbitrary in the constitutional sense,’ ” but the standard is “‘no calibrated yard stick.’” Id. (quoting County of Sacramento v. Lewis, 523 U.S. 833, 846, 118 S.Ct. 1708, 140 L.Ed.2d 1043 (1998)). In situations where, as here, there is an opportunity for the state officer to reflect and to deliberate before acting, deliberate indifference is the appropriate standard. McQueen, 433 F.3d at 469. “We have equated deliberate indifference with subjective recklessness, which means that the official must both be aware of facts from which the inference' could be drawn that a substantial risk of serious harm exists, and he must also draw the inference.” Id. (citations and quotation marks omitted). Subjective recklessness can “be proven circumstantially by evidence showing that the risk was so obvious *466that the official had to have known about it.” Bukowski v. City of Akron, 326 F.3d 702, 710 (6th Cir.2003) (citing Ewolski, 287 F.3d at 513 n. 7).
It is a close call whether Lowery acted with the requisite culpability when he released the unredaeted Internal Affairs investigative report to the inmates. Viewing the evidence in the light most favorable to the nonmoving party, however, as we must when reviewing a summary judgment motion, Brown v. Earthboard Sports USA, Inc., 481 F.3d 901, 902 (6th Cir.2007), I believe factual issues exist concerning whether the risk was so obvious that Lowery had to have known about it.
Specifically, issues remain as to Lowery’s level of recklessness. In an affidavit, Lowery claimed that he was unaware of corrections officers’ information appearing in the report, and that he would not have released it to the inmates had he known. In deposition testimony, however, Lowery admitted to reviewing every page of the report before its release. Accordingly, Lowery’s failure to notice a single corrections officer’s personal information, which appeared on multiple pages of the report, seems suspect. Indeed, one corrections officer’s personal information — including his name, birth date, and social security number — appeared on the very same page, prominently set apart from the paragraphs of text, as a prison informant’s information that Lowery did redact.
Accordingly, I believe that this matter should proceed to a jury to resolve whether Lowery’s awareness of the personal information contained within the report and his failure to redact that information amounted to deliberate indifference and not mere negligence.
II.
Because the facts here are materially indistinguishable from Kallstrom, I respectfully dissent from Part III of the majority’s opinion. Given the striking similarities between Kallstrom and the instant case, I have little trouble concluding that the corrections officers’ constitutional rights were violated and that such rights were clearly established. As for the Plaintiffs’ state-created-danger claim, I would REVERSE the district court’s grant of summary judgment in favor of Lowery and REMAND for further proceedings on whether Lowery acted with deliberate indifference.