08 Jul 2026
When the morning brief landed, it carried a pattern that’s become impossible to ignore. Entra passkey enrollment vishing targets Microsoft 365 users. It deserves more than a passing glance. Because this is not a theoretical risk — it is happening now.
Here is the story in full — not just the headline.
Plenty of outlets will tell you a breach happened. Fewer will tell you what to do with that knowledge. That is what this piece aims to fix.
Entra passkey enrollment vishing targets Microsoft 365 users
Before dismissing this as another breach story, look closer. Entra passkey enrollment vishing targets Microsoft 365 users was reported by BleepingComputer.
The surface-level explanation only tells part of the story. Digging deeper reveals patterns that repeat across incident after incident.
What made this attack effective
- Target reconnaissance: The attacker knew the environment well enough to avoid noisy mistakes.
- Abuse of trust: Legitimate credentials, signed software, or trusted vendor access blurred detection.
- Signal suppression: Logs tampered with, alerts tuned out, or SIEM blind spots where the actor operated.
- Delayed disclosure: The gap between compromise and public knowledge often stretches months.
Attackers do not reinvent the wheel with every breach. They repeat what works because organisations keep making the same mistakes. That is not a failure of intelligence — it is a failure of process.
What this means for the industry
Most cybersecurity coverage reads like a press release. “An incident may have occurred. The company is investigating.” That helps nobody.
What is often missing from the conversation is the human layer. The CFO who disables MFA to save ten seconds. The developer who hardcodes credentials because it is faster. The server that everyone knows is outdated but nobody owns. This is where incidents are born.
The organisations that survive are the ones willing to see their own weaknesses clearly. Pretending the perimeter is fine does not make it so.
New HalluSquatting Attack Could Trick AI Coding Assistants Into Installing Botnet Malware
While that story unfolded, another pattern emerged. New HalluSquatting Attack Could Trick AI Coding Assistants Into Installing Botnet Malware, reported by The Hacker News.
Each story like this is a data point. Collect enough of them and the picture becomes harder to ignore.
Three recurring themes seem relevant here:
- Trust exploitation: Attackers do not break encryption — they break the trust placed in people, processes, or systems.
- Speed over scrutiny: The pressure to ship, deploy, or publish often overrides the time needed to verify.
- Posture drift: Defences are often strong at implementation and weak at maintenance. What was true in January is no longer true in May.
State IDs for AI Agents: Will Estonia Set a Precedent?
While that story unfolded, another pattern emerged. State IDs for AI Agents: Will Estonia Set a Precedent?, reported by Dark Reading. The world’s digital testing ground plans to help people use AI agents for government purposes.
It is easy to dismiss a single headline. The danger is in missing the trend that connects it to everything else.
The uncomfortable truth is that most of these incidents share a common origin: a small decision that seemed harmless at the time. A skipped review. A delayed patch. A credential shared for convenience. Individual moments, but they stack up.
The question is not whether attackers are getting smarter. It is whether defenders are getting complacent. If your security posture has not been materially improved in the last six months, it has probably degraded — because the threat landscape certainly has not stood still.
What ties these stories together
Treated separately, each breach is a headline. Together, they are a trend. Attacks are getting quieter, more targeted, and more patient. The high-profile ransomware events still grab headlines, but the real damage is often done silently — data exfiltrated over months, privileges escalated quietly, backdoors left for later.
The gap between knowing and doing is where most incidents start. Awareness is not protection. Action is.
This is not about fear. It is about honest assessment. The organisations that handle incidents well are not necessarily the ones with the biggest budgets. They are the ones that prepared before they needed to.
Turning awareness into action
Enough analysis. Here is what actually moves the needle. Not the generic advice — the specific actions that reduce risk in measurable ways.
This week
- Audit privileged accounts. Who holds admin rights? When was the list last reviewed? If you cannot answer within thirty seconds, that is a finding.
- Push MFA everywhere. No exceptions. Executive convenience is not a justification for single-factor access.
- Patch public-facing assets first. VPN, gateway, web server — if it touches the internet and it is not current, it is a priority.
- Restore a backup. Time it. If it takes more than two hours, your backup strategy is aspirational, not operational.
- Review logging coverage. Authentication, DNS, file access, privilege use. If any of those is unlogged, detection is blind.
Medium-term improvements
- Segment your network. If one compromised endpoint can reach your domain controller, your segmentation is inadequate.
- Operationalise EDR alerts. Alerts without response are noise. Define who acts, how quickly, and under what conditions.
- Run phishing simulations. Then deliver targeted training. Measure click-rate reduction over time.
- Review third-party access. Vendors, contractors, integrations — if the access is not actively needed, revoke it.
- Update your IR playbook. Make it usable at 3 AM. Role cards, contact trees, decision trees. Not a PDF nobody reads.
Cybersecurity is not a product, it is a practice. And like any practice, discipline matters more than inspiration.
Where this leaves us
Each of these stories carries the same underlying message: the attack surface keeps growing, and the defenders are still adjusting.
The organisations that survive the next wave will be the ones that treat visibility as a discipline, not a product.
There is no silver bullet. But there is absolutely a difference between trying and hoping. Choose the former.
Stay sharp. Stay questioning. And I will see you at the next brief.
