Everpure positions storage as the last line of cyber defense as AI accelerates attacks

Everpure pitches storage as the last line of cyber defense in the AI era - SiliconANGLE

Everpure pitches storage as the last line of cyber defense in the AI era Everpure Inc. today reinforced its Enterprise Data Cloud vision by defining the storage layer as the last line of defense in modern cyber resilience, an architecture the company argues is essential as artificial intelligence accelerates the pace and sophistication of attacks against enterprises. The data management company, formerly known as Pure Storage until it changed its name in February, said its approach assumes that perimeter security will fail. It shifts the focus to guaranteeing that recovery points themselves cannot be corrupted, tampered with or locked out, even by an attacker who has obtained administrative access. Sitting underneath all of this is a control plane that is walled off from the production environment and covers both on-premises and cloud storage. It keeps a clean copy of the data that no production administrator can touch. The point, Everpure says, is to cover the obvious case of an attack wiping data out and the less obvious case of data being quietly corrupted and the damage only showing up months later. By the company's own figures, fewer than three in 10 ransomware victims ever fully recover their files. The architecture is built around three pillars. Trusted Recovery layers in a human-in-the-loop mandate that requires multiparty, out-of-band authorization for sensitive data actions, enforced through Everpure Fusion and a set of Security Presets designed to eliminate configuration drift and keep SafeMode snapshots active by default. Autonomous Resilience uses the Everpure Protect service to correlate external threat intelligence with storage-level telemetry and trigger preemptive hardening. Economic Predictability is delivered through the Evergreen//One subscription model, which the company says removes the trade-off between paying a ransom and absorbing extended downtime as the average cost of a data breach climbs to $4.44 million. Everpure pointed to a recent incident at an unnamed Fortune 100 customer as evidence the architecture works in production. Attackers used stolen credentials and native tools to delete thousands of endpoints and virtual clusters, but Everpure said its strict administrative separation prevented them from touching SafeMode snapshots, allowing the organization to restore revenue-critical operations in hours. The company is also tying recovery to data context through its recently closed acquisition of 1touch.io Inc., whose data discovery technology will be integrated into the Enterprise Data Cloud to map relationships between business applications and the underlying data they depend on. Everpure said the integration will allow customers to prioritize restoration of the most critical workloads first. "The modern enterprise is defined by its data, yet most organizations are flying blind, treating their most valuable asset as a commodity to be warehoused," said Prakash Darji, general manager of digital experience at Everpure. "We are giving our customers the certainty that no matter what happens at the perimeter, the heartbeat of their business stays strong."

Everpure - 'the perimeter has already failed and in a post-Mythos world, recovery speed is everything'

When Anthropic announced Mythos Preview in April and simultaneously said the model was too dangerous to release to the public, the enterprise technology industry - and buyers - went into a spiral. The reason was that Anthropic said that Mythos had demonstrated the ability to find and exploit zero-day vulnerabilities in real open source codebases - and could do so autonomously, overnight, with no formal security training required on the part of the person prompting it. One AI model, in seven weeks, found more than 2,000 previously unknown software vulnerabilities - in software that security researchers had studied for years. Anthropic limited Mythos to a small group of American companies, including Apple, Amazon, JPMorgan Chase and Palo Alto Networks, to reduce the risk that bad actors get their hands on it. Anthropic CEO Dario Amodei has since warned that Chinese AI models are "maybe six to 12 months" behind Mythos, giving organizations roughly that window to fix what Mythos has found. This is the context in which Everpure, the company formerly known as Pure Storage, is putting its Enterprise Data Cloud cyber resilience capabilities front and center ahead of its annual Accelerate event in Las Vegas next month. The company is not announcing new products today - Patrick Smith, Everpure's Field CTO for EMEA, was open about that when we spoke last week. But what it is doing is making an argument that the perimeter in the enterprise has already failed and that the storage layer plays a critical role in protecting organizations in a world where cyber resilience is getting even harder. Smith's view is not that Mythos and AI more broadly have created an entirely new security paradigm, but that it has dramatically compressed the timeline of an existing one - and that compression is forcing organizations to invert long-held assumptions. On the current customer conversations, Smith said: Cyber resilience has been a big ticket item for enterprises for really the last five years, but it continues to accelerate. Just when you think it can't be more on people's minds, more on the agendas of boards - not just the CSO - it goes to a whole new level. On Mythos specifically, Smith pointed to the scepticism around not releasing it to the public, but highlighted that the end result has been greater concern amongst buyers: The trigger for that new level has been Anthropic's release of Mythos. It's fascinating, because to release something and then say this is too dangerous to release to the general public could be said to be the ultimate in marketing. But equally, it's forcing people to take a really accelerated view of the whole cycle - from vulnerability identified, to fix released, to fix deployed across an environment, to vulnerability eliminated. That cycle typically takes a month or two, depending on severity. Now, with the power of AI being applied in this space, we expect that window to be down to days. For years, there has been a lag of weeks to months between patch release and implementation - Mythos does not fix that lag, but it exposes it as critical, as there is a model capable of finding and exploiting it at a speed that makes the conventional patching window extremely high risk. The practical implication, Smith argues, is that the whole approach to patching has to flip: The key thing is velocity - [enterprises] need to operate at a higher velocity than they ever have before in terms of deploying patches. Typically, you deploy patches in a phased approach - so that if you break things, you break the things that don't matter first and avoid breaking the critical ones. With the velocity that could be required in a post-Mythos world, you don't have time to do that. And in fact, you probably want to patch the most critical systems first, because you want to remediate the vulnerabilities that exist on them. So you completely turn the approach on its head. If the capabilities of Mythos prove legitimate, this is a big operational shift for large enterprises. Smith said that patching ten or a hundred servers is manageable. However, patching 50,000 or 100,000 servers - at speed, with critical systems first, accepting that you will break things - is a different category of problem entirely. Everpure has been making the case for immutable snapshots - point-in-time copies of data that cannot be altered or deleted, even by an administrator - and Safe Mode recovery, a hardened configuration that locks down storage arrays against unauthorized changes, for the better part of three years. That argument was always sound, but uptake hasn't always been consistent. Smith points to a changing recognition amongst security teams about their role in protecting against vulnerabilities: There was a time, probably two years ago, when we still didn't have as much uptake of snapshots and our Safe Mode capability as we felt customers should, because storage teams just viewed it as not being their responsibility. Now it's very unusual to find one of our customers who doesn't have it turned on. What Mythos has done, however, is give the snapshot argument a second job. Smith said that in a world where patches need to be deployed at speed, you will break things. That means you need a verified, recoverable snapshot of your environment before you patch - not only to recover from an attack, but to roll back from the disruption your own remediation creates. It's the same capability, but it's addressing a somewhat different problem. Everpure's Enterprise Data Cloud platform delivers this through what it describes as an isolated, intelligent control plane that governs data across on-premises and cloud environments. The company's position is that even an attacker who gains administrative access cannot corrupt, tamper with, or delete protected copies. Everpure pointed to a Fortune 100 organization that suffered a sophisticated malware-free attack - attackers using stolen credentials and native tools to delete thousands of endpoints and virtual clusters. And the vendor says that its capabilities meant that the data layer remained untouched and revenue-critical operations were restored in hours, not weeks. The closing of the 1touch acquisition - announced alongside the rebrand in February and confirmed complete on 11 May - adds an additional dimension to the resilience story that Everpure is pushing here. 1touch is a data intelligence and orchestration company whose platform discovers, classifies, and contextualizes data across all environments - on-premises, cloud, and SaaS. The premise is that recovery speed is only half the equation, and that enterprises also need to know what you have, where it lives, and which of it matters most before an incident occurs. In other words, whilst speed of recovery matters, so does knowing what, and of what, in what order. Smith's said: You can't treat all of your business applications and all of your data as equals. You've got to prioritize. We were discussing this at a user group yesterday - understanding your minimum viable company, the things you bring back first, what everything else can wait for. How do you determine what that looks like? I suggested to Smith that this could become even more complicated as the user experience gets even more extracted away from the data layer thanks to agentic and generative AI - and as those systems interact with data in new ways. Smith said: It'll be interesting over time - and I don't really know what I think about this yet - whether that's going to get easier or harder as data gets further abstracted from the experience of people working with it through AI. The complexity of the data and application landscape - on-premises infrastructure, one or more public clouds, SaaS services that many organizations consume without fully understanding where or how they run - is already significant. The 1touch integration is designed to map the links between business applications and their underlying data so that when recovery is required, the most critical operations are restored first. My guess is that the actual value of that capability will only become clearer as agentic AI embeds itself into more workflows and makes the dependency graph harder to trace manually. Smith said that the business case for the type of cyber resilience architecture that Everpure is discussing is getting easier to make internally. On board-level attention Smith said: I spoke to the CTO of one of our customers last year, who had just secured funding to build out essentially a data bunker for his organization. He'd had a meeting with the CFO and explained that he couldn't guarantee he could recover the business if they didn't have that capability - and funding didn't take long to secure after that conversation. That framing - recovery guarantee as a prerequisite for business continuity funding - will likely resonate well with a CIO/CTO audience. The average cost of a data breach reaching $4.44 million is the figure Everpure cites in its materials and the reputational damage is arguably a harder number to put in a spreadsheet. With more and more companies ending up in the headlines due to cyber incidents, boards and enterprise leaders are increasingly willing to open cash coffers to protect themselves.

Everpure Redefines Cyber Resilience with Data Management as the Last Line of Defense

In the age of AI, storage and data management determine business survival when the perimeter fails Everpure the company revolutionizing storage and data management, today reinforced its Enterprise Data Cloud vision, defining the storage layer as the last line of defense in modern cyber resilience. As AI weaponizes zero-day vulnerabilities and automates sophisticated attacks, CISOs are no longer fighting at human speed. Everpure is leading the transition to an 'outside-in' security model that assumes perimeter failure and guarantees the storage layer remains an uncompromisable point of recovery. By using immutable snapshots, the Everpure architecture transforms restoration into an instantaneous experience, eliminating the manual processes that traditionally inflate Mean Time to Recover (MTTR). To prevent an attacker or rogue AI from weaponizing this speed, Everpure places humans at the governance gate. Recovery is immediate; data destruction is impossible without verified oversight. "The modern enterprise is defined by its data, yet most organizations are flying blind, treating their most valuable asset as a commodity to be warehoused," said Prakash Darji, General Manager of Digital Experience at Everpure. "We are doubling down on that reality. By architecting our platform to be both data-aware and inherently resilient, we aren't just managing data -- we are delivering an insurance policy against the chaos of the AI era. We are giving our customers the certainty that no matter what happens at the perimeter, the heartbeat of their business stays strong." Resilience by Design Everpure's Enterprise Data Cloud delivers guaranteed data recovery, turning days of downtime into minutes. An isolated, intelligent control plane governs data across on-premises and cloud environments and preserves the integrity of the recovery point itself, so even an attacker who gains administrative access cannot corrupt, tamper with, or delete protected copies. Everpure keeps a verified, immutable version of the truth available at all times, protecting customers against both immediate data loss and the long-tail risk of silent corruption that can surface months after an attack- countering the 72% of ransomware-affected organizations that never fully recover their data. This architecture enables three core outcomes: * Autonomous Resilience: Everpure ensures continuous operations through upgrades, patching, and active attacks. This resilience is driven by the Everpure Protect service, which correlates external threat signals with storage-level insights to trigger preemptive hardening. Working in tandem with Everpure Fusion, it acts as an automated active defense layer -- remediating configuration gaps and enforcing security standards across every endpoint to keep an active defender at peak efficacy. * Trusted Recovery: Everpure integrates a Human-in-the-Loop (HITL) mandate, requiring multi-party, out-of-band authorization for any sensitive data action. Enforced by Everpure Fusion, this governance uses Security Presets to eliminate configuration drift and ensure SafeMode snapshots are active by default. Even if an adversary or rogue AI gains administrative control of the production environment, the data layer remains isolated and ready for verified restoration. * Economic Predictability: As the average cost of a data breach reaches $4.44 million[2], Everpure eliminates the ransom vs. recovery gamble. Through Evergreen//One, customers access predictable, subscription-based economics that align infrastructure costs to usage while eliminating the financial sting of disruptive upgrades and downtime. Proven Recovery in Hours Instead of Weeks Detection without recovery is a liability. And when prevention fails, recovery speed is the ultimate benchmark of resilience -- but speed only matters if it is built on a foundation of precision and verified trust. A Fortune 100 organization recently proved this during a sophisticated malware-free attack that bypassed traditional defenses. Using stolen credentials and native tools, the attackers crippled the identity and compute layers, deleting thousands of endpoints and virtual clusters. Despite the devastation, the data layer remained untouchable. Everpure's strict administrative separation prevented attackers from accessing, modifying, or deleting protected SafeMode snapshots -- even with global administrator privileges. This foundation of verified trust allowed the organization to restore revenue-critical operations in hours rather than weeks. Contextual Intelligence Understanding the data itself is the next frontier of defense. With the closing of the 1touch acquisition, Everpure is adding a critical dimension to its resilience pillar: context. Effective recovery depends on understanding data, not just storing it. By integrating 1touch's advanced data discovery with the Enterprise Data Cloud, Everpure customers gain a continuous, 360-degree view of the data landscape. This intelligence maps the vital links between business applications and their underlying data, ensuring that when a recovery is required, the most critical operations are restored with absolute priority and precision. The outside-in approach thrives on integration. Everpure unifies the security landscape by integrating its platform with the threat intelligence, security analytics, and data protection providers. This synchronization ensures that when the perimeter detects a threat, the data layer is an active defender and trusted for immediate recovery.