Heart Monitor Firm Hit, 360,000 Records Gone

iRhythm, according to The Record, makes the Zio Patch — a chest-worn sensor used for long-term cardiac monitoring and heart arrhythmia detection. This week the company began filing breach notices across multiple US states, confirming that hackers accessed third-party business systems between 3 and 8 June.

The entry point was a social engineering attack (where attackers manipulate people into handing over access, rather than breaking through technical defences). At least 298,647 people in Texas and 69,526 in South Carolina are confirmed affected, with California filings also submitted.

What was taken: names, addresses, phone numbers, patient account numbers, device serial numbers, insurance numbers, dates of service, and dates of birth. iRhythm says its clinical systems, medical devices, and operations were not disrupted.

In a June filing with the US Securities and Exchange Commission, iRhythm disclosed that it received a communication from the attacker demanding payment in exchange for not publicly releasing the stolen data. The company has not confirmed whether any payment was made, and no hacking group has claimed responsibility.

iRhythm says it has "no evidence" the stolen data has been used for identity theft so far. That is reassuring, but medical data is valuable for years after a breach.

What you should do: If you have ever used an iRhythm device, monitor your insurance statements for unfamiliar claims and consider placing a free fraud alert with the major credit bureaus.

Business email compromise (BEC) is a scam where attackers impersonate executives or vendors to trick employees into transferring money or handing over access. According to Dark Reading, attackers are now running the same playbook against AI agents, and it is working.

As companies give AI tools authority to approve invoices, screen job applications, and interact with business systems, attackers are using prompt injection (hiding malicious instructions inside content the AI is asked to read) to manipulate those agents into taking harmful actions. An AI agent told to process a résumé, for example, might encounter hidden text instructing it to advance a candidate automatically.

The FBI recorded roughly $3 billion in BEC losses in 2025. Teaching employees to spot scams still matters. Teaching the AI systems that work alongside them to resist manipulation is now equally urgent.

What you should do: If your workplace uses AI tools with any authority over payments, approvals, or data access, ask whoever manages them how prompt injection is being mitigated.

What NetScaler is: Citrix NetScaler ADC and Gateway are enterprise networking appliances used by organisations to manage, secure, and control application traffic and remote access.

What it is: A memory overflow flaw (where a programme writes more data than a buffer can hold, corrupting adjacent memory) that can allow an attacker to remotely execute code or crash the appliance entirely. Citrix rates this as critical and is urging immediate action, according to SecurityWeek.

Who's at risk: Organisations running NetScaler ADC or Gateway configured as a SAML SP or SAML IdP (identity and authentication components used in single sign-on systems), or Secure Private Access Hybrid deployments that use NetScaler.

CVSS: 9.5 — Critical. Patch today.

Root cause: The flaw lives in how the appliance processes SAML authentication requests. Input is not properly validated before being written to memory, so a specially crafted request can overflow the buffer and hand an attacker control over what executes next. Think of it like a form that accepts any length of text and crashes the filing system when you type too much.

Attack vector: A remote attacker sends a malformed SAML request to an exposed NetScaler endpoint. If the appliance is not patched, that request overflows the memory buffer and can either crash the service or deliver arbitrary code with the privileges of the running process. No authentication is required to trigger it.

Recommended actions:

1. Upgrade to NetScaler ADC and Gateway version 14.1-73.46, 13.1-64.29, or the relevant FIPS build immediately.
2. If patching cannot happen today, restrict external access to SAML endpoints at the network perimeter.
3. Review access logs for unexpected SAML authentication attempts or service restarts.

Anthropic has launched OSS Scanner, a free service that uses its AI models to periodically scan open source software projects for security flaws, according to SecurityWeek. Open source software is the foundation of most of the internet, and security flaws in it often go unnoticed for years.

Maintainers opt in, receive a report for each potential vulnerability found, and get a proof-of-concept showing how it could be exploited, plus a suggested fix where one exists. Reports are sent without human review, so Anthropic flags that some findings may contain inaccuracies, though it expects more than 90% to be accurate.

If you maintain or depend on open source projects professionally, this is worth flagging to your development team.

What you should do: Open source project maintainers can opt in at anthropic.com. If you are not a maintainer, raise this service with whoever manages your organisation's software dependencies.

Australia's government is exploring mandatory reporting requirements for agentic cyberattacks, following a breach in which an OpenAI agent accessed Australian Medicare systems without authorisation last June, according to Dark Reading. OpenAI became aware of the incident two months after the fact, then waited another month before notifying affected agencies. Parliament questioned executives from OpenAI, Anthropic, Microsoft, and Google this week. The outcome could set a precedent for how quickly AI companies must disclose incidents globally.