Snare Insider Issue #15 – Oct 2026

Newsletter Issue #15 – Special Cyber Security Awareness Month Edition

The Threat Window Is Shrinking. Is Your Visibility Keeping Up?

October is Cybersecurity Awareness Month, and this year the Snare Insider is doing something different. Instead of one monthly issue, we are building the full Snare Insider over four weeks, one instalment at a time. This issue will continue to grow each week, so make sure you check back in each week.

Awareness is the starting point, but the teams we talk to want something more practical: what is changing, what it means for their environment, and one thing they can do about it this week.

Here is how the month unfolds.

Week 1 sets the scene with the Global Cyber Threat Pulse and a Threat Spotlight on the investigation gap.

Week 2 brings three feature articles timed with InfoSec World in Orlando and CyberCon in Melbourne.

Week 3 introduces the Snare perspective, the next evolution of Snare, a webinar and a practical playbook.

Week 4 rounds out the month.

Every instalment keeps the same thread: visibility before, during and after an incident matters more than ever. This week we start with the threat environment itself, because the case for better evidence begins with how quickly it is changing.

  • Attackers are moving faster. 88% of exploitation CrowdStrike observed against vulnerabilities with a public proof-of-concept in the first half of 2026 happened within 48 hours of release.
  • Identity is the attack path. Vishing intrusions doubled and device-code phishing attempts rose 15x in the same period — and ENISA’s 2026 Threat Landscape names social engineering and vulnerability exploitation as the two leading ways attackers get in across the EU.
  • Your historical evidence still matters. Global median dwell time rose to 14 days in Mandiant’s M-Trends 2026, and only 52% of compromises were first identified internally. Keeping the evidence is step one; being able to ask it questions fast is the next constraint.

The common thread: visibility before, during and after an incident matters more than ever.

TL:DR

1.Global Cyber Threat Pulse October 2026

Four regions. Four headline incidents, plus the smaller stories sitting alongside them. One recurring problem: organisations need independent, retained, searchable evidence.

AUSTRALIA — A THIRD-PARTY VULNERABILITY, NOT QUEST’S OWN SYSTEMS

Quest Apartment Hotels’ first disclosure described “primarily names, email addresses, and/or other contact details.” A mid-September update, after forensic analysis, confirmed about 1.99 million customers were affected, including 297,739 card numbers (46,727 with CVV), 104,268 passport and/or driver’s licence numbers and a small number of Medicare and NDIS numbers. The entry point was a vulnerability at a third-party service provider, not Quest’s own systems.

ASIA — THREE MONTHS BETWEEN ACCESS AND DISCLOSURE

A compromised maintenance-employee account gave attackers access to Japan’s Government Solution Service files in late June. A July investigation traced the entry point to an exploited VPN vulnerability, and the digital agency disclosed in September that more than 246,000 records were compromised. The sequence is the story: access in June, root cause in July, public disclosure in September.

UK/EUROPE — THE EMAIL PASSED EVERY TECHNICAL CHECK

Revolut confirmed it released customer identity data to an unauthorised third party after a fraudulent information request arrived from a real government agency’s domain and passed SPF, DKIM and DMARC. No system intrusion was involved: domain authenticity and requester authority are two different things.

US/NORTH AMERICA — THE API HAD NO RATE LIMIT

A threat actor claimed to have extracted 7.49 million customer records from a CenterPoint Energy public API said to lack rate limiting and adequate authentication. CenterPoint confirmed the breach in an SEC Form 8-K on 14 September, and class-action filings allege the exfiltration ran from 17 August to 1 September.

Threat Spotlight 

2.The Investigation Gap

This month we keep coming back to evidence: what gets collected, what gets retained, what’s still there when an investigation starts. That’s the right starting point — but it isn’t the whole problem. A second, quieter constraint sits right behind it: the investigation gap, the widening distance between how much security teams collect and how much of it is genuinely askable when it matters.

What the 2026 research says:

  • 79% of SOCs already use AI or machine learning somewhere in their operations; only 36% have integrated it into a defined SOC workflow (SANS 2026 SOC Survey, June 2026).
  • 24% of cyber leaders name enterprise-wide visibility as the single biggest barrier to SOC effectiveness — the top-ranked answer (SANS 2026 SOC Survey).
  • 70% of large SOCs are expected to be piloting AI agents for Tier 1 and Tier 2 operations by 2028, but only 15% will achieve measurable improvement without structured evaluation first (Gartner, 2026).

Read together, the picture is integration and context. Teams already have plenty of alerts and tools; what makes the difference is evidence integrated enough to ask a question across the whole environment at once. Retention answers “does the evidence still exist?” Integration answers “can someone find and use it, correlated, in the time an investigation has?” Four ordinary questions show where that matters:

  • “Has this account done this anywhere else in the last 90 days?”
  • “Did we see this indicator anywhere before the alert fired?”
  • “What changed on this host immediately before and after the event?”
  • “Is this pattern happening anywhere else in the environment right now?”

None of these are exotic — they’re the questions an analyst asks in the first ten minutes. In most environments, each still has to be hand-translated into platform-specific syntax, source by source, before it can even be asked.

Closing the gap takes three things at once: evidence that is collected and retained; evidence that is correlated and queryable in something closer to plain language than six query syntaxes; and, per Gartner’s caution, a querying layer that sits on evidence that’s already integrated. That third piece is where the Snare Insider goes later in the month, with a first full look at AskSnare, built on an evidence layer that’s already collected, normalised and correlated. Not a replacement for analyst judgement, and not the bolted-on AI SOC agent Gartner is cautioning against.

Source: SANS 2026 SOC Survey, June 2026; Gartner, “Validate the Promises of AI SOC Agents With These Key Questions,” 2026

Feature

Week 1 – Do This, This Week

Spend ten minutes pressure-testing your logging environment. Ask three questions:

  • Are we collecting the evidence we’d need to reconstruct an attack? Think beyond endpoint logs to identity, authentication, cloud platforms, network infrastructure, applications and privileged activity.
  • Are we retaining it long enough? Detection may happen today — the first sign of compromise may have occurred weeks or months earlier.
  • Could we actually retrieve and investigate it quickly? Collecting logs only has value if your team can access, search and analyse them fast once an incident occurs.

Week 1 Resources

The Cybersecurity Visibility Readiness Checklist

A practical checklist for security teams to assess:

Collection → Coverage → Integrity → Routing → Retention → Storage → Investigation → Incident Recovery

Use it to identify where gaps in your logging architecture could become gaps in your incident investigation.

Keep reading:

  • The 10 Questions Every CISO Should Be Able to Answer From Their Logs — a practical test of whether your organisation has the visibility required to investigate suspicious activity.
  • The Ransomware Investigation Playbook — what to log, retain and query at every stage of a ransomware intrusion, from initial access to extortion.

Feature Article

Can You Prove What Your AI Did?

The AI conversation in cybersecurity is changing.

The question is no longer simply whether security teams should use AI.

It is increasingly:

What authority should an AI system have — and how do you prove what it did with that authority?

That question appears repeatedly across this week’s CyberCon and InfoSec World programs.

At CyberCon, sessions are examining how AI outputs can be made contestable, logged and defensible under audit, how enterprises govern autonomous systems, and whether organisations can demonstrate that an AI agent made the correct decision.

InfoSec World is tackling the same issue from another direction. Its program describes AI agents with many of the characteristics traditionally associated with privileged users: identities, credentials, access to sensitive information, connections into enterprise applications and the authority to execute actions.

That has an important consequence for security architecture.

AI activity needs to become observable activity.

Consider an AI-enabled security system that recommends or initiates a response.

Six months later, could you establish:

  • which identity initiated the interaction?
  • what data the AI was able to access?
  • which logs or evidence informed the decision?
  • what applications, APIs or tools it invoked?
  • whether permissions changed during the process?
  • what action was recommended?
  • what action was actually executed?
  • whether a human approved, modified or rejected it?
  • what happened immediately afterwards?

If those questions cannot be answered, the problem isn’t simply AI explainability.

It is an evidence gap.

Logging needs to follow the action

Traditional audit logging typically focuses on human users, applications, operating systems and infrastructure.

Agentic systems introduce another actor.

An AI agent may authenticate, query data, call another service, modify infrastructure or trigger remediation. Those actions need the same characteristics security teams expect from other security-relevant events:

identity, timestamp, source, target, action, outcome and context.

And the logs describing that activity need to be protected and retained independently enough that they can be used to validate what occurred.

This becomes particularly important when AI is being used inside the SOC itself.

InfoSec World’s From Alert to Evidence session highlights the potential for natural-language investigation to accelerate evidence gathering while still requiring authorised scope, transparent actions, evidence-grounded findings, human validation and a defensible record.

That distinction matters.

AI can accelerate an investigation.

It shouldn’t replace the underlying evidence.

The security question for 2026

As organisations move further into autonomous and agentic systems, security teams should add another requirement to AI governance:

Can we reconstruct the actions of the AI system as reliably as we can reconstruct the actions of a privileged administrator?

If the answer is no, logging and auditability need to become part of the AI architecture — not something added after deployment.

Key takeaway

If AI becomes part of the decision chain, it must also become part of the evidence chain.

Feature Article

Detection Is Not Investigation

Security operations have become very good at producing signals.

EDR, XDR, SIEM, identity platforms, cloud security tools and threat intelligence services can identify suspicious behaviour quickly.

But an alert answers a very specific question:

Where should we start looking?

It does not necessarily answer:

What happened?

That distinction appears repeatedly across this week’s conference programs.

InfoSec World’s From Alert to Evidence session describes a familiar security operations problem: a high-confidence alert can be produced within seconds, while determining what occurred, how far activity spread and which systems were affected can still take hours or days.

Another session focuses on moving security operations from simply chasing alerts toward understanding attacks through improved visibility, contextual investigation and higher-fidelity detection.

This is fundamentally an evidence problem.

Start before the alert

Imagine an identity alert triggers at 10:17 Tuesday morning.

The investigation may need to determine:

  • when the credentials were first compromised;
  • which device originally authenticated;
  • whether authentication occurred from another location;
  • what privilege the account had at the time;
  • whether that privilege changed;
  • what systems the identity subsequently accessed;
  • which processes executed on those systems;
  • whether administrative tools were used;
  • whether data was accessed or moved;
  • whether similar activity occurred previously.

The alert at 10:17 may be only one event in a sequence that began days or weeks earlier.

The investigator therefore needs something much broader than the detection itself.

They need a timeline of evidence.

Evidence rarely lives in one place

A modern investigation can require events from:

Endpoints and servers
Authentication, process execution, services, file activity, configuration changes and administrative events.

Identity systems
Logins, authentication failures, privilege changes, MFA activity, service accounts and directory changes.

Cloud infrastructure
Control-plane actions, workload events, configuration changes and access activity.

SaaS platforms
User activity, administrator changes, application permissions and data access.

Network and security infrastructure
Connections, policy changes, firewall activity and security-control events.

The challenge isn’t simply collecting each dataset.

It is being able to correlate them.

An investigator needs to move from:

identity → device → process → destination → action → outcome

without losing the sequence because one dataset has expired, another was never collected, timestamps aren’t aligned or the evidence exists inside a system the investigation team cannot readily search.

Historical evidence matters

NIST’s current incident-response guidance treats incident response as part of broader cybersecurity risk management rather than something that begins only after an incident has been detected. Preparation, detection, response and recovery need to work as an integrated capability.

Logging architecture belongs in that preparation.

The time to discover that an important identity log was retained for only seven days is not on day eight of an investigation.

The time to discover that endpoint evidence exists in one platform while infrastructure evidence sits somewhere else isn’t while trying to determine lateral movement.

And the time to decide whether historical logs are trustworthy isn’t when they are being presented as forensic evidence.

Collection needs to support investigation

That means security teams need to think beyond:

“Are we collecting logs?”

and ask:

“Could we reconstruct an incident from them?”

Those are not the same question.

Collection without sufficient coverage creates blind spots.

Collection without retention creates historical gaps.

Collection without integrity weakens confidence in the evidence.

Collection without searchability delays investigations.

And collection without context leaves analysts trying to manually assemble relationships between thousands or millions of individual events.

Key takeaway

Detection tells you something happened. Evidence lets you determine what happened.

Your logging architecture needs to support both.

Feature Article

Retention Is Not Ingestion

Security teams face an uncomfortable contradiction.

They need more evidence.

But sending more evidence into the SIEM can make security operations significantly more expensive.

Those two requirements are often treated as if they are inseparable.

They aren’t.

Australia’s current practitioner guidance on SIEM and SOAR implementation explicitly warns that many SIEM pricing models are linked to the quantity of logs ingested and that poor control of ingestion can create significant costs. The same guidance recommends defined logging standards and retention policies aligned to organisational risk and regulatory requirements.

Its companion guidance also recommends prioritising logs for SIEM ingestion based on their value to threat detection, investigation and incident response.

The principles are globally relevant.

Three decisions — not one

Logging strategies should distinguish between:

  1. Collection
    What evidence does the organisation need to capture?
  2. Retention
    How long should that evidence remain available?
  3. SIEM ingestion
    Which events require immediate analytics, correlation, detection and alerting?

Those decisions overlap.

But they do not need to produce identical datasets.

Not every retained event needs to be a continuously ingested SIEM event

High-value security events may justify immediate SIEM ingestion because they contribute directly to detections and security monitoring.

Other events may have significant forensic or compliance value but comparatively little value in continuous real-time analytics.

The architecture therefore becomes a question of data placement.

What needs immediate detection?

What needs rapid search?

What needs longer-term forensic retention?

What can be aggregated or de-duplicated?

What can be retained efficiently and replayed into analytics when an investigation requires it?

And what no longer provides sufficient security value to justify keeping at all?

The cost question is really an architecture question

Simply reducing logging is a poor answer to SIEM cost pressure.

Missing evidence can be far more expensive when an investigation occurs.

The better question is:

Does every event need to be stored and processed in the same way?

For many organisations, the answer is no.

A deliberate log architecture can maintain broad evidentiary coverage while controlling what enters the most expensive analytics and storage layers.

That also reduces another architectural risk: making the organisation’s evidence dependent on one SIEM.

Your evidence should survive a SIEM decision

Organisations change SIEMs.

They consolidate vendors.

They move workloads.

They adopt new cloud platforms.

They change MSSPs.

They introduce new analytics and AI capabilities.

The historical evidence required for security investigations should not disappear simply because the analytics platform changes.

That makes the collection layer strategically important.

Logs should be capable of being collected, normalised, retained and routed independently of the destination that happens to analyse them today.

This provides security teams with another option when requirements change:

route the evidence somewhere else.

Key takeaway

Retaining security evidence and continuously ingesting it into a SIEM are different decisions.

Treating them separately can improve both investigative readiness and cost control.

Week 2- Do This, This Week

Run an Evidence Gap Drill

You don’t need another maturity assessment to identify your most important logging gaps.

Run an investigation.

Not a full incident-response exercise — just a 30-minute evidence test.

Step 1 — Pick a scenario

For example:

A privileged account is suspected of compromise today. Initial access may have occurred 60 days ago.

Step 2 — Pick your evidence

Can you retrieve the relevant:

  • authentication events?
  • identity and privilege changes?
  • endpoint and server activity?
  • cloud control-plane events?
  • SaaS administrative activity?
  • security-control events?
  • network activity?

Step 3 — Test the timeline

Attempt to reconstruct:

identity → device → access → privilege → action → destination → outcome

Step 4 — Record where the investigation stops

Look specifically for:

  • evidence that was never collected;
  • data that has already expired;
  • information that exists but isn’t searchable;
  • evidence held in a platform investigators cannot easily access;
  • inconsistent timestamps;
  • missing identity context;
  • gaps between endpoint and cloud/SaaS activity;
  • evidence available only through the current SIEM.

Step 5 — Ask one final question

If this were a real incident, what would we be unable to prove?

That is your evidence gap.

Week 2 Resources

Security Investigation & Auditability Checklist

AI agents now hold identities, credentials and the authority to act. This checklist helps security and governance teams test whether those actions can be reconstructed when it matters. Its 40 checks across eight areas cover AI inventory and identity, through to action logging, human oversight and retention. A 90-day test then shows where your evidence gap is.

Feature

Attackers Follow Trust, Not Infrastructure

Exploited vulnerabilities and phishing now open more breaches than stolen credentials, but identity is still how intrusions spread. ENISA lists valid accounts among the core techniques of the ransomware groups it tracked in 2025, and Microsoft found 52.2% of valid-account intrusions led to further credential theft.

Valid accounts, OAuth grants and SaaS sessions give attackers direct routes to data, often with no malware and no lateral movement to observe.

We take the investigation chain from Week 2’s drill and map where each link now sits across identity, SaaS, cloud and host evidence.

We also look at what Microsoft’s critical infrastructure data and Gartner’s agent prediction mean for the evidence you keep.

Feature

One Incident. Many Evidence Planes.

ENISA names attacks on SaaS, CRM and cloud environments among the major trends of 2025, and NSA and CISA rank cloud identity management among their top cloud mitigations. Microsoft found cloud identity abuse was the most observed technique against critical infrastructure.

Week 2 established that evidence rarely lives in one place.

This article covers what it takes to join it: the identity and session keys that link one plane to the next, the data-plane logs that aren’t on by default, why every investigation is limited by its shortest retention period, and the evidence that sits with your suppliers rather than with you.

Week 3- Do This, This Week

Last week’s drill tested the evidence you hold. This week, test the evidence someone else holds for you.

Choose your most critical SaaS or identity provider (the one an attacker would most want to reach), then:

  1. Find out what exists. Which audit logs does the provider generate for your tenant, and at which licence tier?
  2. Find out how long it is kept. What is the default retention, and what does it cost to extend it?
  3. Request it. Export 30 days of administrative and authentication activity for one privileged account, through the API or console, as you would during an incident.
  4. Time it. How long until you had the first record? How long until you had all 30 days?
  5. Test the join. Do its timestamps, user identifiers and source addresses line up with your own identity and endpoint evidence?
  6. Ask about the rest. What activity does the provider log on its side that you can’t export yourself, and how would you get it during an incident?

Anything you couldn’t get, couldn’t get quickly or couldn’t join is part of your evidence gap, now outside your own boundary.

Week 3 Resources

The Security Evidence Coverage Map

Map where investigation evidence exists across your environment, and beyond it. For each source, capture:

Evidence source Collected? Retention Searchable? SIEM? Owner Gap
Endpoint
Server
Identity
SaaS
Cloud
Network
Security platforms
Third parties
AI / agent activity

The objective isn’t to get every box into the SIEM. It is to know what evidence exists, where it exists, how long it exists and whether an investigator can use it.

Feature

THE LONG-DWELL PROBLEM

Six incidents. One uncomfortable question.

Six recent breaches across five countries went undetected for between 85 days and 20 months. In each one, the outcome depended on what evidence still existed when the investigation finally began.

  • Discovery comes late. IBM’s 2026 data puts the average breach at 183 days to identify and 247 days to identify and contain. Breaches running past 200 days cost an average of USD 1.33 million more.
  • The first sign is rarely a security alert. These breaches surfaced through a different investigation, a ransomware event, a supplier notice and an IT performance problem. At South Staffordshire Water, only 5% of the IT environment was being monitored.
  • Lost logs become part of the incident. About 13% of Coupang’s logs covering the attack period were lost, partly through deletion after regulators had ordered them preserved. As a result, not every victim could be identified.

The question: if you found a breach today, could your evidence take you back to where it began?

The Snare Perspective

Don’t build another log pipeline. Build an evidence layer.

Snare doesn’t replace the SIEM. It keeps your security evidence independent of the SIEM, so detection can keep evolving while the evidence trail stays complete, trustworthy and usable for as long as an investigation needs it.

  • Collect and protect at the source. Snare Agent v5.11 captures telemetry beyond the Event Log, including ETW providers such as DNS Analytic logs. An encrypted local buffer keeps events through outages, and sequence-numbered streams make any gaps visible.
  • Retain more than you send. Snare Central keeps evidence from 90 days to 7+ years, well beyond the 247-day average breach lifecycle, in hash-verified archives. Snare Reflector routes only what each SIEM needs, to several destinations at once.
  • Replay and investigate when the question changes. Event Replay sends a historical window back into Microsoft Sentinel or Splunk. AskSnare, a cybersecurity AI assistant, queries retained evidence where it lives and explains its reasoning, while the analyst decides.

The test: ask any logging platform to demonstrate the ten capabilities in our technical test, not just claim them.

THE TECHNICAL TEST

Ten questions to ask any logging platform

Marketing language makes most security platforms sound alike. Architecture shows the differences. Before choosing a collection or log management platform, ask it to demonstrate, not just claim, each of these:

# Ask it to demonstrate How Snare answers
1 Collection of forensic evidence directly from the OS, including telemetry beyond the Event Log Snare Agent v5.11 collects ETW providers, including DNS Analytic and Debug logs
2 Collection that continues when the SIEM or network is unavailable, with gaps you can detect Encrypted local disk buffer; sequence-numbered streams
3 Full historical retention independent of SIEM licensing and ingestion Snare Central: 90 days to 7+ years on local, SAN or S3-compatible storage
4 Filtering, aggregation and de-duplication of SIEM traffic without destroying the investigation record Filtering applies to what is forwarded; aggregation keeps original detail; the complete record is retained
5 Simultaneous delivery to multiple SIEMs, archives and data platforms Reflector: independent queues per destination, with failover
6 A SIEM migration without rebuilding endpoint collection Parallel delivery to old and new SIEMs from the same agents
7 Replay of a historical window into analytics for a new investigation Event Replay into Microsoft Sentinel and Splunk
8 Investigation of retained evidence without copying it to another platform first Central search; AskSnare queries data in place
9 AI findings you can trace back to the evidence behind them AskSnare gives a rationale with every insight; the analyst decides
10 A complete architecture that runs fully on-premises Pathway 1, with no dependency on cloud or AI

If several answers are no, you don’t have an independent security evidence layer. You have a forwarding pipeline.

Week 4- Do This, This Week

Run the 247-Day Recovery Test

Week 2’s drill asked where your investigation would stop. This one tests whether your historical evidence is usable: whether it can be retrieved, replayed and trusted.

Assume you discover a compromised privileged account today, and that the breach followed IBM’s average: initial access 247 days ago. Choose that identity and, for each window, test four things:

Window Retrievable within an hour? Searchable without a restore project? Replayable into your analytics? Provably unaltered?
90 days ago
183 days ago (average time to identify)
247 days ago (average lifecycle)

Then ask:

  • Which sources expired before 247 days, and was that a decision or a default?
  • Which events were filtered out before they were retained, not just before they reached the SIEM?
  • Which evidence exists only inside the current SIEM?
  • Would your logging pipeline itself show a gap, such as an agent that went quiet or a log that was cleared?

If this were a real incident, which part of the evidence would you fail to produce or defend? That’s the logging gap to fix first.

Week 4 Resources

The Intelligent Log Management Playbook

This month’s incidents share one lesson: the evidence you need is decided long before the incident. The Intelligent Log Management Playbook turns that into a plan, from collection through to AI-assisted investigation, so the evidence is there when discovery comes late.

Inside:

  • The Security Evidence Chain: eight links from event to action (collect, protect, normalise, optimise, route, retain, investigate, act) and the design decision behind each.
  • Five decisions before technology: what to collect, what enters the SIEM, what to retain, where data can reside and how you will investigate.
  • Ten use cases, from compromised credentials and lateral movement to ransomware reconstruction, historical threat hunting, SIEM migration and missing telemetry. Each one lists the evidence required and example questions to ask of it.
  • Hot, warm and historical data tiers, separating what the SIEM must analyse now from what investigations need months later.
  • A five-level maturity model, an architecture checklist and a six-step assessment to work through with your security, infrastructure and risk teams.

OCTOBER RESOURCE PACK

Four weeks. Four practical tools.

Cybersecurity Awareness Month is ending, but these resources are built to stay useful.

  • Cybersecurity Visibility Readiness Checklist: assess collection, coverage, integrity, routing, retention, storage and investigation readiness. [Download the checklist]
  • AI Security Investigation & Auditability Checklist: test whether AI-driven decisions and actions can be independently reconstructed and verified. [Download the checklist]
  • Security Evidence Coverage Map: map where endpoint, server, identity, SaaS, cloud, third-party and AI evidence exists, and where the gaps are. [Download the map]
  • Intelligent Log Management Playbook (new this week): a deliberate approach to collection, retention, SIEM ingestion, cost control and forensic investigation. [Download the playbook]

Cyber Security Awareness Week Wrap up

FROM AWARENESS TO EVIDENCE

We started October with a simple idea: awareness is useful; visibility is actionable. Over four weeks that took us from visibility to evidence:

  • what you collect;
  • whether you can prove what AI did;
  • where security evidence now lives;
  • and whether you can still reconstruct an attack when discovery comes months later.

SIEMs will change. Infrastructure will move. AI will become part of every investigation. Threat actors will keep operating through legitimate identities, trusted systems and connected services.

One requirement stays constant. Keep the evidence. Keep it trustworthy. Keep control of where it goes. And make sure investigators can use it when the question finally arrives.

Collect. Govern. Investigate. Defend.

Sources

This issue draws on primary regulatory text and named industry benchmarking rather than secondary summaries. Key sources:

CrowdStrike, “2026 Threat Hunting Report,” August 2026. crowdstrike.com/en-us/blog/crowdstrike-2026-threat-hunting-report/

ENISA, “Threat Landscape 2026,” September 2026. enisa.europa.eu/topics/cyber-threats/threat-landscape

Mandiant / Google Threat Intelligence Group, “M-Trends 2026,” 2026. cloud.google.com/blog/topics/threat-intelligence/m-trends-2026

SANS Institute, “2026 SOC Survey,” June 2026.

Gartner, “Validate the Promises of AI SOC Agents With These Key Questions,” Craig Lawson and Andrew Davies, 2026.

ASD’s ACSC, CISA et al., Implementing SIEM and SOAR Platforms: Practitioner Guidance, May 2025, Section 4 (Principles 4 and 7): https://www.cyber.gov.au/sites/default/files/2025-05/implementing_siem_and_soar_platforms_-_practitioner_guidance.pdf

ASD’s ACSC, CISA et al., Priority Logs for SIEM Ingestion: Practitioner Guidance, May 2025: https://media.defense.gov/2025/May/27/2003722069/-1/-1/0/PRIORITY-LOGS-FOR-SIEM-INGESTION-PRACTITIONER-GUIDANCE.PDF

PCI Security Standards Council, Payment Card Industry Data Security Standard (PCI DSS) v4.0.1, June 2024, Requirement 10.5.1: https://www.pcisecuritystandards.org/document_library/

Mandiant (Google Cloud), M-Trends 2026, 23 March 2026: https://cloud.google.com/blog/topics/threat-intelligence/m-trends-2026

IBM Security, Cost of a Data Breach Report 2025. Reported in Help Net Security, 4 August 2025: https://www.helpnetsecurity.com/2025/08/04/ibm-cost-data-breach-report-2025/

CISA, “Guidance for SIEM and SOAR Implementation” (series overview): https://www.cisa.gov/resources-tools/resources/guidance-siem-and-soar-implementation

Gartner, “Gartner Predicts Over 40% of Agentic AI Projects Will Be Canceled by End of 2027,” press release, 25 June 2025. Reported in SD Times: https://sdtimes.com/agentic-ai/gartner-more-than-40-of-agentic-ai-projects-will-be-canceled-in-next-few-years/

CyberCon 2026 Program, “Umpiring AI decisions: The Decision Review System (DRS) framework for contestable and board-ready security governance,” Wednesday 14 October 2026: https://melbourne2026.cyberconference.com.au/program

InfoSec World 2026, “When AI Agents Go Live: Who Controls Their Access?”: https://www.infosecworldusa.com/isw26/session/?iswid=4449866

InfoSec World 2026, “Invisible Administrators: Non-Human Identities as the New Enterprise Attack Surface”: https://www.infosecworldusa.com/isw26/session/?iswid=4121361

InfoSec World 2026, “From Hallucinations to Controls: Designing Auditable AI Systems in the Real World”: https://www.infosecworldusa.com/isw26/session/?iswid=4119042

OWASP GenAI Security Project, Agentic AI — Threats and Mitigations (threats T3, T8, T9). Summary: https://humansecurity.com/learn/blog/agentic-ai-security-owasp-threats

IBM Security, Cost of a Data Breach Report 2025. Reported in Help Net Security, 4 August 2025: https://www.helpnetsecurity.com/2025/08/04/ibm-cost-data-breach-report-2025/

Regulation (EU) 2024/1689 (Artificial Intelligence Act), Articles 12 and 19: https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-19

ISO/IEC 42001:2023, Information technology — Artificial intelligence — Management system, Annex A, control A.6.2.8 “AI system recording of event logs.”

InfoSec World 2026, “From Alert to Evidence: Investigating Enterprise Threats in Natural Language,” Tech Theater, 14 October 2026: https://www.infosecworldusa.com/isw26/session/?iswid=4443726

Snare Solutions
Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.