Debunking Common Myths about Cyber-Security Culture
A useful myth-buster replaces a misleading claim with a better explanation, not an equally sweeping slogan. These entries distinguish research findings from my interpretation and recommendations. The purpose is to improve organisational decisions, not to blame colleagues or promise that any single intervention will eliminate risk.
The better explanation: This label hides the system around the person. The NCSC culture principles focus on leadership, shared expectations and conditions that enable secure behaviour. Its phishing guidance also warns against excessive reliance on people spotting attacks.
I would ask which controls failed, what demands competed and whether the approved action was practical. People still have responsibilities, but so do those designing and funding the work. Calling a colleague a weak link tells us remarkably little about how to improve anything. Give people workable tools and support, then examine the whole chain rather than selecting the most convenient link to blame.
The better explanation: Completion establishes participation in the recorded activity. It does not show whether someone can verify a payment change, use the sharing controls or obtain urgent help. The article makes that distinction for good reason.
NIST’s 2024 guidance incorporates evaluation and improvement throughout a learning programme’s lifecycle. My recommendation is to combine reach with supported practice, workflow evidence and response information. Investigate whether excluded groups could access the material before drawing conclusions about coverage. A full register is useful administration; it is not evidence that every relevant decision will be handled safely. Keep the register, but do not ask it to answer a different question.
The better explanation: A scheduled course can contribute, but it cannot anticipate every role change, new service or revised process. The ICO’s training guidance calls for regular assessment of staff groups’ needs and a relevant programme, rather than treating delivery as an end in itself.
I would combine a common foundation with task-specific practice and updates when work changes. That does not mean bombarding people with weekly reminders. Choose the intervention because a decision or capability needs support, and provide working time for it. A calendar is useful for arranging delivery; it is not a substitute for deciding what colleagues actually need to learn.
The better explanation: Fear is not a repair strategy. The NCSC phishing guidance notes that stress and situational pressures affect clicking, and that threatening punishment does not change those factors. It also warns that fear can undermine reporting.
My recommendation is to explain a relevant risk, provide a practical action and make help accessible. In a hypothetical payment-change briefing, rehearse verification and escalation rather than presenting a parade of alarming headlines. Threat information can be useful when it informs a decision. It is less useful when it leaves somebody anxious, busy and no better equipped. We should be supplying tools, not simply adding dramatic background music.
The better explanation: A click does not tell you why it happened or which intervention would help. The NCSC warns against blame and punishment in phishing simulations, including their effects on trust and willingness to report.
I would use a concerning result to examine the task, message, workload and available support. Offer relevant practice where there is a genuine learning need, and keep any established misconduct process separate. That is not the same as abandoning accountability. Accountability should reach the people responsible for confusing controls and contradictory targets too. A training exercise should illuminate a problem, not become a disciplinary ambush with an educational logo.
The better explanation: Helping someone report promptly and deciding how conduct should be addressed are different tasks. The NCSC culture principles support safety, trust and openness; they do not require a blanket promise that every action is acceptable.
I recommend clear reporting support, careful fact-finding and fair, established procedures for any separate conduct question. Be honest about how information may be used. Do not promise absolute immunity, then quietly withdraw it. Equally, do not make the first response to disclosure an interrogation about blame. We need the facts early enough to act, while preserving both organisational responsibility and fair treatment of the person involved.
The better explanation: The message, audience, filtering and working circumstances may all have changed. The NIST Phish Scale rates human detection difficulty, highlighting why the email itself belongs in the interpretation.
I would state the counting rules, coverage and scenario differences, then ask what other evidence supports a learning effect. A lower rate is an observation, not automatic proof of cause or reduced future losses. Compare like with like where possible and describe the limitations where that is not possible. If several controls changed together, report the combined improvement rather than awarding all the credit to the course. The neatest explanation is not necessarily the one the evidence supports.
The better explanation: More reports could reflect increased exposure, a simulation, improved access to reporting or greater willingness to ask for help. The number alone cannot distinguish those explanations. In Lain and colleagues’ 2022 company study, collective reporting showed operational value.
I would examine what was reported, how quickly it reached responders, whether the route worked and what action followed. Conversely, silence is not proof that everything is fine. Explain the population and reporting opportunities behind any rate. Treat reporting as part of the response system, not a nuisance statistic to drive down until the dashboard looks peaceful.
The better explanation: The need is useful learning and workable defences, not a particular exercise format. The NCSC recommends layered phishing mitigation and discusses significant limitations of simulations.
I would choose a simulation only when it answers a defined question that a less intrusive exercise cannot answer as well. A tabletop can explore decisions; supported practice can test a workflow; checking the reporting route can reveal an operational gap. These are recommendations for choosing methods, not claims that one format always wins. The burden of justification belongs with the proposed intervention. “Other companies do it” is a reason to ask questions, not a procurement requirement.
The better explanation: Studies of one training method cannot settle the value of every kind of learning. The 2022 Lain study tested particular embedded phishing interventions at one company. The 2024 study identified reminder effects rather than benefits from consuming the embedded content in its setting.
Those findings challenge a common practice, not the entire idea of developing capability. I would distinguish recognising scam messages from practising access administration or coordinating an incident. Examine the question, setting, intervention and outcome before generalising. Evidence-led opinion means being willing to change your view; it does not mean replacing a supplier’s universal promise with an equally universal dismissal.
The better explanation: IT cannot independently change sales incentives, allocate every team’s working time or decide how a director responds to challenge. The NCSC principles explicitly emphasise leadership and the wider conditions supporting behaviour.
I would give IT clear responsibility for its services and controls while keeping business owners accountable for their workflows. HR, learning, privacy and security each contribute their expertise. Shared work still needs named decisions and escalation routes. Otherwise, “send it to IT” becomes a remarkably efficient way of moving a management problem without solving it. Culture belongs in operational and leadership decisions, not solely in the technical support queue.
The better explanation: Senior leaders make decisions about resources, exceptions and disruption, so their learning should reflect those responsibilities. The NCSC culture principles give leaders a vital role in setting organisational expectations.
I would ask executives to practise decisions under incomplete information, including customer communication and authority to pause a service. They should also follow the controls they expect others to use. This does not require transforming directors into engineers. It requires informed judgement and a willingness to ask questions. The most expensive misunderstanding in a meeting may be the one nobody challenges because it came from the end of the table with the largest job title.
The better explanation: Consistent objectives do not require identical delivery or content. The ICO’s guidance calls for assessment of different groups’ training needs and responsibilities. The article also sets out why equivalent access matters.
I recommend a common foundation with practice matched to tasks, access and decisions. Provide adjustments and alternatives so people can achieve the objective. Do not infer learning needs from age, ancestry, gender or job-title stereotypes. At 61, I am particularly unpersuaded by the idea that a birth date is a diagnostic tool. Ask what the person needs to do and what support makes that possible. That is a more useful basis for both fairness and programme design.
The better explanation: The relevant questions are what work someone performs and what information or systems they can access. The ICO’s training-needs guidance explicitly includes voluntary, temporary and contract staff.
I would provide essential local instructions before the relevant task and check access to help and reporting. Coordinate responsibilities with the provider or employer where appropriate, but do not assume their generic course explains your arrangements. Keep the delivery proportionate to the role and duration. A short engagement is a reason to make instructions clear and timely, not a reason to postpone them until the person has already left.
The better explanation: Acknowledgement and usability are different things. The NCSC’s principles-in-practice scenarios illustrate how restrictions without workable alternatives can push teams towards unofficial file-sharing tools.
I recommend walking through the approved process with the people expected to use it. Check whether the contact record exists, the required tool is available and urgent exceptions have an owner. A signature may document receipt; it cannot mend a broken reporting route. Where workarounds appear, investigate their purpose before assuming indifference. The policy should explain a workable arrangement, not serve as a receipt transferring responsibility for an impractical one.
The better explanation: Additional collection needs a purpose and justification, not merely a storage location. The ICO’s worker-monitoring guidance warns that excessive monitoring can intrude into private lives and affect rights and wellbeing.
I would specify the decision being supported, collect proportionately and restrict access and retention. Check whether small groups remain identifiable even in an aggregate report. Avoid individual league tables or covert reuse for productivity monitoring. Treat privacy advice as part of programme design, not the final obstacle before launch. People deserve an honest account of what is collected about them and why. A dashboard’s appetite is not a lawful basis or a business need.
The better explanation: As the article explains, the legal requirement for appropriate technical and organisational security measures is broader than completing a course. It should not be reduced to a universal product specification or assumed schedule.
The ICO’s all-staff training toolkit describes a strategic, needs-based programme and regular assessment of relevant groups. It is guidance and is currently under review. I recommend checking current legal, regulatory, sector and contractual requirements with the appropriate owner, then documenting how the programme addresses them. Do not infer that training is optional, either. The point is to meet the actual obligations, rather than an oversimplified version conveniently shaped like a supplier’s catalogue.
The better explanation: A platform can support delivery and records; it cannot decide whether a manager permits a payment delay or funds a reporting service. The NCSC principles place leadership and organisational conditions at the centre of cultural improvement.
I would define the need, test the proposed support and include internal capacity in the budget. Ask who maintains content, handles adjustments, receives reports and interprets results. If those responsibilities are unresolved, purchasing more functionality does not resolve them. Technology may be useful, but culture is not something a vendor can install while everyone else carries on unchanged. Pilot the service against a real workflow before making a long commitment.
The better explanation: The article’s roadmap is a recommended first cycle, not evidence of a transformation deadline. NIST’s learning-programme guidance calls for continuing evaluation and updates as needs evolve.
I would review what changed, which groups were included, what remained difficult and whether the evidence justifies expansion. Keep ownership for maintaining resources as systems and roles change. Pause where reporting, support or privacy arrangements remain unresolved. Success might mean discovering an obstacle early and repairing it, not achieving an unqualified green status. A pilot earns the next decision; it does not grant permanent exemption from asking whether the arrangements still work.
The better explanation: The conflict may lie in the design of a particular process, not in the goals themselves. The NCSC’s illustrative sales scenario shows security and sales teams agreeing a shared purpose and co-designing workable procedures. It is an example of the approach, not a quantified outcome study.
I recommend making verification, escalation and explanations part of the service design. Give colleagues authority to challenge unusual requests while offering customers a legitimate way forward. Some safeguards will still introduce necessary limits or delay; explain the reason and manage the hand-off. Customers and staff are people to support, not obstacles to control. That is the organising principle I would keep when the next fashionable security slogan arrives.
Putting the FAQs and myth-busters to work
Use these FAQs and myth-busters as conversation starters, not another set of answers for colleagues to memorise. Choose a question or claim that connects with work your team actually does, such as checking a supplier’s payment details, sharing information or reporting a mistake. Ask whether the explanation matches your arrangements, where the approved process becomes difficult and what support is missing. Nobody needs to recite the article. I would settle for a useful discussion followed by something being fixed.
For onboarding or a team briefing, pair a relevant FAQ with a myth that challenges an unhelpful assumption, then practise the approved response using non-sensitive material. Managers can use the same approach to examine their own instructions and priorities. Invite colleagues to question the advice and describe exceptions, without asking anyone to disclose personal mistakes in front of the group. Keep the focus on improving the work, not identifying somebody to blame. If the discussion reveals an unclear rule, inaccessible tool or unsupported reporting route, assign it to an owner with the authority to act.
For programme reviews and purchasing decisions, use the entries to frame questions about relevance, evidence, accessibility, privacy and internal capacity. Check the linked sources and current requirements before incorporating an answer into policy or training, particularly where legal duties are involved. Distinguish a research finding from my interpretation, and adapt recommendations to your circumstances rather than treating them as universal instructions. Finish with one agreed improvement, an accountable owner and a review date. Then return to the relevant question and ask what changed. The point is not to win an argument with a myth; it is to make secure working more practical for colleagues and customers.
