Application security testing in Baltimore.
Application security testing in Baltimore, Maryland means testing the systems that Baltimore organisations actually run — defense and government contracting and healthcare and health systems estates — against Maryland's breach-notification clock of no later than 45 days, and against MODPA. Every engagement also grades your TLS A–F for quantum exposure.
Why Baltimore organisations commission application security testing
Baltimore is a federal cybersecurity and intelligence corridor adjacent to major academic medical and research institutions. That economic profile decides what an attacker goes after here, and it should decide what a test is scoped to look for. A generic methodology applied without reference to the local industry mix produces a report full of findings nobody needed and silence on the ones that mattered.
What is specific to this market
The federal cybersecurity corridor means an unusually informed buyer and an unusually high bar. Clients frequently employ people who have done offensive work at national scale, so credibility rests on methodology and evidence quality. Maryland’s privacy act imposes strict data-minimisation duties that go beyond the Virginia template, and its notification rule requires notifying the AG before consumers.
Defense and government contracting
Contractors and subcontractors are attacked as the soft route to information the prime is defending properly. Controlled Unclassified Information sitting in an engineering share, a supplier portal or an email archive is the objective, and the adversary is patient, well-resourced and specifically interested in persistence rather than immediate extraction.
CMMC assessment turns what used to be a self-attestation into an audited control set, and flow-down clauses mean the requirement reaches suppliers who have never been assessed before.
Healthcare and health systems
Healthcare is the sector where a security failure becomes a clinical failure. Ransomware against a hospital does not merely encrypt files; it diverts ambulances and delays procedures. The pressure to restore service fast is precisely what makes healthcare a profitable target, and the sprawl of clinical devices, imaging systems and third-party portals gives an attacker unusually many ways in.
The HIPAA Security Rule requires periodic technical evaluation, and payers and health systems increasingly demand HITRUST or an equivalent third-party assessment before they will integrate a vendor.
Biotech and life sciences
The asset is intellectual property with a decade-long development cycle: assay data, trial results, manufacturing process detail. The adversary is frequently seeking a research advantage rather than a payday, which changes the profile — the goal is quiet persistence and exfiltration, not disruption you would notice.
Validated GxP environments constrain how and when testing can run, so scheduling and change-control planning matter as much as methodology.
What Maryland law expects of you
Security testing is not a compliance exercise, but in Maryland the legal clock is what turns an unnoticed weakness into a reportable event with a deadline attached. Knowing the timeline in advance is what lets you decide how fast findings need to be remediated.
| Obligation | Requirement in Maryland |
|---|---|
| Consumer notification deadline | no later than 45 days |
| Regulator notification | AG must be notified before consumers are. |
| Comprehensive privacy statute | Maryland Online Data Privacy Act (MODPA) — in effect |
MODPA imposes strict data-minimisation duties that go beyond the Virginia template most states copied.
The practical consequence for Baltimore businesses is straightforward: a breach you discover on a Friday starts a clock that runs in calendar days, not business days. Testing exists to find the exposure before that clock ever starts — and to give you documented evidence of diligence if it does.
How the engagement runs
Application security testing means driving your deployed application the way a motivated attacker would: authenticating properly, walking the flows a user walks, and then deliberately doing the things a user is not supposed to be able to do. It catches the class of bug that only exists at runtime — broken access control, injection reachable through a specific parameter, session handling that survives a logout, a rate limit that is not actually applied.
Coverage is the whole game. An unauthenticated crawl of a single-page app finds a login form and nothing else. Secuur configures authentication first — session cookies, bearer tokens, OAuth flows, whatever your app uses — then imports your OpenAPI or GraphQL schema so the API surface behind the front end is exercised too.
1. Configure & authenticate
We set up credentials, roles and API schemas, then verify coverage with you before the first real pass.
2. Test
Automated passes across the full surface, with a human reviewing anything that touches business logic or authorisation.
3. Retest
You fix, we re-run the specific proof for each finding and mark it closed with evidence. Retests are included.
Working with Baltimore teams
Engagements run remotely by default, which keeps scheduling simple and cost down. Where a scope genuinely needs physical presence — internal network testing from inside a facility, an operational technology environment, a physical access assessment — on-site time is planned into the statement of work rather than billed as a surprise. Reporting, walkthroughs and remediation support run on your calendar, not ours.
What the engagement covers
| Area | What we do |
|---|---|
| Authenticated coverage | Session, JWT, OAuth 2.0 and multi-step logins, with re-authentication when a session drops mid-scan. |
| API-first testing | Import OpenAPI, Swagger, Postman collections or a GraphQL schema and test every operation, not just the ones the UI happens to call. |
| Business-logic abuse | IDOR, privilege escalation between roles, price and quantity tampering, workflow steps executed out of order. |
| Injection & OWASP Top 10 | SQL, command, template and header injection; SSRF; XSS in every reflected and stored context; unsafe deserialisation. |
| SPA & JS-heavy apps | A real browser engine renders the app so client-side routes and XHR endpoints are discovered rather than missed. |
| Reproducible evidence | Every finding ships with the exact request, the response that proved it, and a curl one-liner your developer can paste. |
What these engagements typically surface
Application testing consistently produces its most serious findings in authorisation logic rather than in the injection classes that dominate training material. Authorisation is application-specific, which is exactly why automated tools cannot evaluate it.
Broken object-level authorisation
An endpoint that checks whether you are authenticated but not whether the record you requested belongs to you. Changing an identifier in a request returns another tenant’s data. This remains the single most common serious finding in modern application testing.
Broken function-level authorisation
Administrative operations exposed to ordinary user roles because the check lives in the interface rather than in the API, so calling the endpoint directly bypasses it entirely.
Tenant isolation failures
Multi-tenant platforms where a scoping filter is missing on one query path among hundreds, or where a background job processes records without the tenant context the request path enforces.
None of the above is hypothetical or specific to Baltimore — these are the recurring patterns across engagements of this type. What varies by market is which of them carries the most consequence, and in Baltimore that is shaped by defense and government contracting and healthcare and health systems exposure more than by anything else.
How to prepare
Provide credentials for every distinct role, including at least two accounts in separate tenants. Cross-tenant testing is where the highest-severity findings come from, and it is impossible without a second account.
In Baltimore specifically, the framing that produces the most useful engagement follows from the local picture: the federal cybersecurity corridor means an unusually informed buyer and an unusually high bar. Clients frequently employ people who have done offensive work at national scale, so credibility rests on methodology and evidence quality. Scope the work against that, not against a generic checklist.
Define the question, not just the scope
The most valuable engagements start from a business question rather than an asset list. "Could an attacker reach our defense and government contracting data from an ordinary employee laptop?" gives testers an objective and gives you a report you can act on. A scope that says only "test everything" produces breadth at the cost of the depth that actually changes decisions.
Fix the cheap findings first
If you already know a system is unpatched or a service is exposed, remediate it before testing begins. Paying senior testers to rediscover issues you have already identified spends the engagement budget on confirmation rather than discovery.
Agree the remediation path before the report lands
Decide in advance who receives findings, who assigns them, and what the target remediation window is by severity. In Maryland this matters concretely: the breach-notification clock runs no later than 45 days, so the difference between a finding fixed in a week and one that sits in a backlog for a quarter is the difference between a managed risk and a reportable event.
Plan for the retest
A finding is not closed until it has been verified closed. Build the retest into the engagement rather than treating it as a separate purchase, and keep the evidence — it is what an auditor, an enterprise customer or a cyber insurer will ask to see.
Compliance drivers that apply in Baltimore
These are the frameworks that most often make testing a requirement rather than a choice for organisations in this market. Which ones bind you depends on your sector, your customers and the data you hold.
- CMMC Level 2
- NIST SP 800-171
- DFARS 252.204-7012
- FedRAMP (cloud services)
- CNSA 2.0
- HIPAA Security Rule
- HITRUST CSF
- FDA premarket cybersecurity (device makers)
- SOC 2 Type II
- 42 CFR Part 2 (behavioural health)
- 21 CFR Part 11
- GxP validation
- HIPAA (clinical data)
- NIST SP 800-171 (federal grants)
- Maryland breach notification — consumer notice no later than 45 days
- Maryland Online Data Privacy Act (MODPA)
The layer nobody else tests
Your application assumes TLS made the channel confidential. That assumption is what "Harvest Now, Decrypt Later" attacks: a session recorded today against a classical-only key exchange is a plaintext session on the day a quantum computer arrives — every token, every payload, retroactively.
CNSA 2.0 sets a 2030–2035 timeline for national security systems to complete post-quantum migration, and that requirement flows down through contract terms. Defense suppliers are the one segment with a hard, published deadline — post-quantum readiness here is a contract-eligibility question.
Every Secuur engagement grades each TLS endpoint in scope A–F on the key-exchange group it actually negotiates, using the same engine as our free readiness scan. Grade A means a hybrid post-quantum group such as X25519MLKEM768; a classical-only handshake grades D to F, because a session recorded today can be decrypted once a cryptographically relevant quantum computer exists. See the NIST post-quantum standards for the underlying algorithms, or the glossary for the terminology.
What you receive
- Findings report with severity and reproduction steps
- Per-role authorisation matrix results
- API endpoint coverage report
- Raw request/response evidence for every issue
- Per-host A–F quantum-readiness grade
- Free retest of remediated findings
- Per-endpoint A–F post-quantum readiness grade
- Attestation letter suitable for customers and auditors
Frequently asked questions
What is the difference between DAST and SAST?
SAST analyses source code without running it and finds patterns like unsanitised inputs. DAST tests the deployed, running application and finds what is actually exploitable — including whole classes such as broken access control that source analysis cannot see. Mature programmes run both.
Can you test APIs with no user interface?
Yes, and this is often where the real risk sits. Import an OpenAPI, Swagger or Postman definition, or a GraphQL schema, and every operation is exercised directly with valid authentication.
Will testing break our production data?
We test against staging by default. Where production testing is required, we agree a written scope, exclude destructive operations, use dedicated test accounts and throttle request rates to stay within your capacity.
How do you avoid false positives?
Findings are proof-driven: a result is only reported when the scan captured a response that demonstrates the behaviour. Anything involving business logic or authorisation is reviewed by a human before it reaches your report.
Do you provide application security testing in Baltimore?
Yes. Secuur serves Baltimore and the surrounding Maryland market. Engagements run remotely by default, with on-site time scoped in where the work genuinely requires physical presence — internal network testing, operational technology environments or physical access assessments.
How quickly must we report a breach in Maryland?
Maryland requires consumer notification no later than 45 days. AG must be notified before consumers are. Those timelines run from discovery or determination, so the practical window to investigate and prepare notice is much shorter than the headline number suggests.
What does an engagement cost?
Scope drives price. A tightly scoped single-application or external test typically starts in the mid four figures; a multi-system engagement covering defense and government contracting infrastructure runs materially higher. We scope from your actual estate — the free readiness scan is usually the fastest way to establish what that estate looks like.