Chapter 1 · Lesson 1.3
The stakes: assets, impacts and orders of magnitude
Cybersecurity is not a purely technical subject: it protects assets that keep the organisation running, and the consequences of an attack affect production, finances, reputation and the law.
The targets: four types of assets
files and databases, procedures, manuals, archives.
software packages, custom software, operating systems, development tools.
servers, workstations, laptops, network equipment, telephone exchange, room air conditioning.
power supply, general services, production lines, manufacturing processes.
The cost of damage varies greatly depending on the asset hit: losing a workstation does not have the same consequences as losing the computer that controls a production line.
The final impacts
| Type of impact | Examples |
|---|---|
| Production | shutdown or loss of ability to produce goods or services: technical malfunction, denial of service, data loss |
| Direct financial | altered invoices, processing errors, fraudulent transfer orders, ransom, cost of restoration |
| Competitive | theft of customer data or know-how |
| Image and reputation | loss of trust of customers and partners; the incident is often publicly known |
| Legal and regulatory | liability in case of fraudulent use of customer data or company resources; breach of obligations |
Orders of magnitude
Updated: October 2026. Sources: CESIN barometer (11th edition, 2026), IBM “Cost of a Data Breach 2026” report (Ponemon Institute).
- Frequency in France: according to the 2026 CESIN barometer (a survey of security managers of large companies), 40% of the organisations surveyed report having suffered at least one significant cyberattack in 2025, against 47% for 2024. The barometer counts only attacks that had a real impact (activity, data, compliance, image); among the companies hit, 81% report an impact on their activity.
- Cost of a data breach: the 2026 IBM/Ponemon study (published late July 2026, on breaches that occurred between March 2025 and February 2026) gives a worldwide average cost of USD 4.99 million per breach, a record, up about 12% from USD 4.44 million in the 2025 edition (which was itself down 9% from 2024). The average time to identify and contain a breach reaches 247 days.
These figures come from surveys by private bodies (user clubs, vendors, consultancies) on particular samples: they give orders of magnitude, not exact measurements of “the” population of companies. They are not comparable from one source to another. The statistics of incidents handled by ANSSI are of a different nature: they describe what is reported to or handled by the agency.
The economics of attack
Cybercrime has organised itself into a service economy: phishing kits, malware “for rent”, already obtained access to company networks, dedicated hosting, resale of stolen data. Attackers with little technical skill can thus launch a campaign at a modest cost, while a poorly protected organisation can lose weeks of activity. Asymmetry is the heart of the problem: the attacker needs to succeed only once, the defender must succeed all the time.
The prices of these services vary over time and between published studies; the figures circulating in the specialist press are rarely verifiable: this course does not reproduce them. Remember the idea: an attack can be rented or bought at low cost, which lowers the skill level required.
A few historical landmarks
| Year | Event |
|---|---|
| 1971 | Creeper, an experimental program that moves from one machine to another on ARPANET (regarded as the first “worm”; the first virus in the strict sense dates from the 1980s) |
| 1988 | Morris worm: first massive spread on the Internet |
| 1989 | “AIDS Trojan”: first known ransomware (distributed on floppy disk) |
| 2000 | ILOVEYOU worm: worldwide spread by email |
Meanwhile, industrial systems adopted the same technologies as office IT (Ethernet, TCP/IP, Windows): the first programmable logic controller dates from 1968 (Modicon) and the Modbus protocol from 1979, designed in a world where malice was not feared. This gap explains a large part of the present fragility of industrial systems (see the “Cybersecurity” course).