Overprotection Breeds Greater Vulnerability | Generated by AI

Home 2026.03

Question: How does excessive emphasis on stability and security in engineering, such as in banks that fear upgrading packages and libraries, ultimately lead to outdated systems and more cybersecurity issues? Is this related to Morita Shoma’s ideas, like people who are overly sensitive about cleanliness ending up becoming dirty?

Answer:

Yes, the phenomenon you describe is a well-known paradox in software engineering, operations, and risk management — especially in highly regulated environments like banks, insurance companies, government systems, and critical infrastructure.

The core pattern is:

This happens through several mechanisms:

  1. Known vulnerabilities accumulate
    • CVEs are published every day.
    • If you freeze dependencies at a 3–5-year-old version, you inherit all publicly known exploits that were fixed in later releases.
  2. The attack surface keeps growing
    • New attack classes appear (Log4Shell, Spring4Shell, MOVEit, text4shell, ProtoPipe, etc.).
    • Old versions usually do not receive patches for brand-new vulnerability types.
  3. Supply-chain and transitive dependency risk increases
    • Even if your direct dependencies look “stable”, their transitive dependencies rot.
    • After 4–7 years almost no library is realistically maintainable without upgrades.
  4. Tooling and observability fall behind
    • Modern security tools (SCA, runtime application self-protection, exploit detection) often refuse to support or give poor results on 5–10-year-old stacks.
  5. Organizational knowledge decays
    • The people who originally built and understood the old stack leave.
    • New engineers treat it as a scary black box → even riskier to touch.

This leads to the ironic outcome:

The organizations that scream loudest about “stability and security” often end up running some of the least secure and least stable long-term environments.

This is very close in spirit to the ideas expressed by Morita Shoma (森田正馬), the founder of Morita therapy (森田療法), even though he was talking about psychology rather than software:

Key parallel statements from Morita therapy:

So in both cases we see the same structure:

Over-protection → constriction → loss of adaptive capacity → worse outcome than moderate risk acceptance

In software this is sometimes called:

Healthy middle path in both psychology (Morita) and engineering:

That is why many security-conscious organizations today (Cloudflare, Google, Stripe, GitHub, etc.) actually upgrade very aggressively — because they understand that stagnation is the bigger risk.

References:


Back Donate