Risk Shield Corp • Updated November 2026
📝 Claims & Coverage • 10 min read

Seven Insurance Mistakes Penetration Testing and Security Consultancies Make

Most security consultancies buy a policy written for generic IT contractors, then sign enterprise contracts it won't honor. Here are seven coverage gaps that hit testers and MSSPs, and what to negotiate instead.

SPONSORED CONTEXTUAL PLACEMENT
💼
Evaluating Enterprise Deployments in 2026?

Request our confidential total cost of ownership matrix and vendor RFP templates.

Launch Calculator →

When an authorized test reads like an intentional attack

If your firm breaks into client networks for a living, the first question about any policy isn't the limit or the premium. It's whether the wording accepts that your core service looks, on paper, exactly like the conduct the insurer excludes. Much of the professional liability insurance for cybersecurity firms on the market is sold on forms built for software developers and general IT consultants. Those forms carry exclusions that sit very close to what a red team does on an ordinary working day.

Mistake: accepting standard intentional-act and unauthorized-access exclusions

Many tech E&O and professional indemnity forms exclude claims arising from deliberate, knowing or intentional acts. Some also exclude claims arising from unauthorized access to computer systems. A tester exploits a vulnerability on purpose. If the client later disputes that a system was in scope, or a third-party hosting provider says it never consented, the insurer has two exclusions to point to before it even looks at negligence.

Advertisement

What to do instead: insist on an affirmative grant of cover. The definition of "professional services" should name penetration testing, vulnerability assessment, red teaming, social engineering, security monitoring and incident response. Narrow the intentional-act exclusion to acts committed with intent to cause harm, and only after a court or arbitrator has made a final finding. Add a severability clause so one rogue employee's conduct can't void cover for the firm and everyone else.

Mistake: assuming out-of-scope activity is covered

Good-faith scope errors are common. A scanner hits a production database that wasn't on the target list. A phishing simulation reaches a client's supplier. A cloud asset turns out to belong to a different tenant. Some carriers attach a condition requiring written authorization for every system tested, and some exclude anything "outside the agreed scope" outright. Read that clause before you buy penetration testing insurance.

What to do instead: treat your rules of engagement as part of the insurance file. Make sure your statement-of-work template, authorization letter and third-party permissions (cloud providers, managed hosts, parent companies) match what the policy condition requires. Ask your broker to confirm in writing that inadvertent, good-faith out-of-scope activity is covered. Computer misuse laws differ by country: the UK's Computer Misuse Act, the US Computer Fraud and Abuse Act, India's IT Act, and equivalent laws across the EU, the Gulf states and Southeast Asia. Criminal defense is rarely covered in full, so check whether the policy carries even a sublimit for it.

RECOMMENDED SOLUTION

Looking to streamline vendor RFP selection?

Access pre-negotiated tier discounts through our independent institutional network.

Compare Pricing →

Uncapped indemnities the policy was never going to pay

Banks, telcos, retailers and government buyers often put unlimited indemnities in their master services agreements, usually for data breach, confidentiality and intellectual property. Security vendors sign them because the deal is big and procurement won't move. The problem is the contractual liability exclusion, which most tech E&O forms contain in some version. It removes cover for liability you've assumed under contract, except where you'd have been liable without the contract anyway.

So the policy pays what you'd owe for ordinary negligence, up to the limit. Anything you promised on top sits on your balance sheet.

Mistake: signing "all losses arising from any security incident" language

Say a 25-person consultancy with $4 million in revenue carries a $2 million tech E&O limit. It signs a retailer's MSA making it liable, without a cap, for "all losses arising from any security incident connected with the services." Its annual test misses an exposed API, and a breach eight months later costs the retailer $6 million. Even if negligence is proven, the policy stops at $2 million. The retailer can also argue the indemnity applies whether or not the tester was negligent, which is the part the contractual liability exclusion strips out. The consultancy could face a claim bigger than its annual revenue.

Fines and penalties make it worse. Whether regulatory fines are insurable depends on where you are, and some EU and UK positions treat certain fines as uninsurable on public-policy grounds. Payment card assessments are often sublimited. Service credits and liquidated damages are usually excluded.

What to do instead:

  • Negotiate a liability cap tied to fees (commonly 12 months' fees or a multiple of contract value) or to your insurance limit, whichever is greater.
  • Keep uncapped carve-outs narrow: fraud, willful misconduct, and possibly confidentiality breach. Breach losses as a whole shouldn't be uncapped.
  • Strike "regardless of fault" wording so the indemnity follows negligence.
  • Ask your carrier for a contractual liability carve-back. Some forms will cover specified indemnities, such as confidentiality or IP, on request.
  • Send every non-standard indemnity to your broker before signature, not at renewal.

Cyber and tech E&O bought as two policies that point at each other

A security firm has two separate exposures. One is third-party claims from clients who say your service failed them, which is tech E&O. The other is your own breach, which needs first-party cyber cover plus privacy liability. The dangerous claims fall between the two. The classic MSSP case is an attacker who compromises your remote management or SOC tooling and pivots into client environments.

Mistake: placing cyber and E&O with different carriers without coordinating them

When two insurers are involved, each can argue the other should respond first. The E&O carrier calls it a security event. The cyber carrier points to its professional-services exclusion. Meanwhile nobody funds your defense. Tech E&O for security consultants works best when the cyber and E&O sections are written together and the definitions mesh.

What to do instead: for most firms with modest revenue, a combined tech E&O and cyber policy from one carrier is the better fit, because it removes the coverage dispute. If you have to split them, align the retroactive dates and the definitions of "claim" and "security event," and get the other-insurance clauses coordinated in writing.

ScenarioSection that should respondCommon gapWhat to confirm
A pen test misses a critical flaw and the client is later breachedTech E&OWarranty or guarantee exclusion triggered by SOW languageNo "failure to detect" carve-out; SOW avoids guarantees
Your SOC or RMM tool is compromised and attackers reach clientsCyber (your breach) and E&O (client claims)Each policy excludes the other's triggerSingle carrier, or a written coordination clause
Ransomware on your own network, no client impactCyber first-partyRansomware sublimits or coinsuranceSublimit size; business interruption waiting period
A tester's laptop holding client data is stolenCyber privacy liability, possibly E&OUnencrypted-device exclusionEncryption warranties match actual practice
A regulator investigates an incident you handled for a clientCyber regulatory or E&O defenseFines uninsurable in some jurisdictionsRegulatory defense sublimit and governing law

Territory clauses and monitoring limits sized for the wrong risk

Mistake: ignoring where claims can be filed

With remote testing, it doesn't matter where your team sits. It matters where a client can sue. A consultancy in London, Bengaluru, Dubai or Singapore may hold a policy with worldwide territory but a jurisdiction clause that only responds to suits brought at home. Some policies exclude US and Canadian jurisdiction outright. If your client is a US subsidiary and the contract is governed by New York law, that exclusion can leave the whole account uninsured.

Local regulation adds another layer. In some countries, including Brazil, India and China, and parts of the Middle East, rules can require locally admitted insurance for local entities. A single global policy may not respond cleanly there, so have your broker check each market.

What to do instead: list your clients by domicile and by the governing law of each contract, then match territory and jurisdiction to that list. Expect adding US jurisdiction to raise the premium, and price it into US engagements rather than dropping it. If you have operating entities abroad, ask about a global program with locally issued policies.

Mistake: sizing MSSP limits to revenue rather than client exposure

MSSP liability insurance usually sits on one annual aggregate shared across every client. Monitoring firms carry a systemic exposure that consultancies don't: one bad detection rule, one faulty agent update or one compromised console can hit dozens of clients at once.

Say an MSSP with 120 clients carries a $5 million aggregate. A misconfigured endpoint policy pushed to every tenant switches off ransomware protection, and 15 clients are hit within a week. If each client's claim runs to several hundred thousand dollars in recovery costs and lost revenue, the $5 million is spread thin across all of them, and defense costs often erode the same limit.

What to do instead: model your worst plausible aggregated event, not your average claim. Check the "related claims" clause. Bundling all client claims into one saves you multiple retentions, but one limit then has to cover everything. Consider excess layers, defense costs outside the limit where available, and contract caps that still make sense when you add up all clients at once.

Retro dates, subcontractors and the renewal file to build now

Mistake: losing your prior-acts cover and forgetting who is insured

Tech E&O is usually written on a claims-made basis. A test you ran three years ago can turn into a claim today, and it's covered only if your retroactive date reaches back that far. Firms lose this when they switch carriers for a cheaper quote and accept a new retro date, or when they're acquired without arranging run-off cover. Multi-year run-off periods are common in some markets, such as UK professional indemnity practice. Others vary.

The definition of "insured" is the other weak point. Many consultancies use freelance testers and specialist subcontractors for hardware, OT or social engineering work. If independent contractors aren't insureds, or if the policy excludes vicarious liability for them, a subcontractor's mistake is your uninsured problem. Directors need attention too. A breach at your own firm can bring claims from investors or customers alleging mismanagement, and that's D&O territory, not E&O.

What to do instead: keep a continuous retro date across every renewal and carrier change. Require subcontractors to carry their own E&O at a stated minimum, and confirm your policy covers your liability for their work. If you have outside investors or are raising money, buy D&O, and arrange run-off before any sale closes.

Indicative pricing for budgeting

These are typical 2026 market ranges for combined tech E&O and cyber. They vary widely by contract terms, controls, claims history, US exposure and region. Confirm with brokers and carriers before budgeting.

Firm profileTypical limitIndicative annual premium (USD)
Boutique or solo tester, under $1M revenue$1M$2,500 to $7,000
Consultancy with 10 to 50 staff, $2M to $10M revenue$2M to $5M$8,000 to $40,000
MSSP or SOC provider, $10M to $50M revenue$5M to $10M plus excess$40,000 to $150,000 or more

Before your next renewal, send a broker who specializes in technology risk three documents: your hardest client contract, your standard rules of engagement, and your list of clients by country. Ask them to show you, in writing, which clause in the proposed wording affirmatively covers penetration testing and monitoring, and how each indemnity in that contract would be treated. The decision rule is simple: if they can't name that clause, take your professional liability insurance for cybersecurity firms to a different market. Final coverage decisions belong with a licensed broker and, for contract terms, qualified legal counsel.

Quick Answers

Does a standard professional indemnity policy cover penetration testing? +
Do security consultancies need both cyber insurance and tech E&O? +
Will our insurer pay if a client sues under an uncapped indemnity we signed? +
How much does penetration testing insurance cost? +
RS
Risk Shield Corp Editorial Team
Independent research desk
Guides on Risk Shield Corp are produced by our editorial team. Pricing, features and regulations change often - confirm details with the vendor or a licensed professional before making decisions.
SPONSORED 2026 Vendor Pricing Benchmarks: Model exact seat pricing, setup fees, and ROI across our verified database.
Launch Calculator →