Every year, the number of cyber attacks is higher than the previous year. The last few years have been no exception, with the Microsoft Digital Defense Report showing a 20% increase worldwide between June 2023 and June 2024.

Far from being spared, France is the 5^th most impacted European country. The preferred targets? Large organizations such as:

And yet, for years now, businesses and institutions have been aware of the problem and investing more in IT security. However, they face more numerous, better organized, more inventive cyber-criminals who use more powerful technologies.

Cyber-criminals have a wide range of ways to access businesses, and all users and businesses can be targeted.

For example, phishing – consisting in sending emails that look like they come from a legitimate organization or familiar company (such as a bank, energy supplier, telecom operator or Social Security) – is designed to trick users into clicking on a link or attachment. Once opened, the malware or ransomware is downloaded and encrypts the files or system access, making it impossible for the user to interact with their computer resources. The cyber-criminals then demand a ransom in exchange for a key to decrypt the data or restore the access.

Very popular with cyber-criminals, phishing and ransomware are highly damaging to businesses. Not only do the ransom and the suspension of activity while the system is shut down have an economic impact on the business, but the risk of data ending up on the dark web (the part of the Internet not indexed by traditional search engines) also damages the business’s reputation.

Another attack technique: SQL injections. In this case, cyber-criminals play on web application vulnerabilities by inserting SQL code into login forms or search bars. The code then interacts with the databases to access critical business information.

Cyber-criminals also target vulnerabilities in information systems by attacking software flaws (especially due to update failures), configuration errors, or security protocol vulnerabilities.

While all these cyber-crime techniques have continued to be perfected over time, the arrival of AI makes them even more sophisticated and difficult to detect.

For phishing, AI’s ability to collect and analyze a target’s data means it can create personalized, highly convincing emails. Spotting the scam is much more difficult for the recipients. 

In the case of vulnerabilities, AI is used to automate the search for software vulnerabilities and bypass traditional defense systems such as antivirus or firewalls.

Finally, AI also makes it possible to automate mass ransomware distribution.

Even if there’s no such thing as zero risk, it can be approached with a pragmatic technological approach and a security culture embraced by all users. To be able to face the threat, it is essential to keep a constant watch on the evolution of cyberattacks. This constant watch makes it possible to anticipate new forms of attack, understand emerging vulnerabilities, and rapidly adapt defense strategies.

These in-house protective measures are essential, but not sufficient. A company’s security also depends on that of its ecosystem. Often connected to sensitive customer data, service providers and subcontractors are a privileged point of entry for hackers. This vigilance is further strengthened by the European NIS2 directive which requires the entire supply chain to adopt appropriate data protection measures and comply with cybersecurity standards.

Raising awareness and providing employees regular training on threats and best practices is a key part of security policy.

For example, users should regularly be reminded of the importance of:

To make sure all these good practices are taken on board by employees, malicious emails produced in-house can be sent. It’s a test that has no impact on the company, but which leaves a lasting impression far more permanent than any speeches. 

Last but not least, organizing regular training sessions not only keeps users on their toes, but also provides up-to-date information on the latest techniques used by cyber-criminals. Without an up-to-date security culture shared by all employees, businesses remain easy prey.

If AI increases cyber-criminal performances, it is also a major ally for cyber defenders. Thanks to its ability to process large volumes of data, AI can:

Using AI to develop cyber-defense tools requires the involvement of experts and training: businesses then gain in defensive reactivity and intrusion prevention.

Faced with the growing sophistication of cyberattacks, especially with the advent of AI, businesses need to adopt a comprehensive cyber-defense approach that combines technology and human factors. A constant watch, the deployment of robust technical solutions, and ongoing employee training have become essential. Security must also extend to the company’s entire ecosystem, including its partners and subcontractors, as required by the NIS2 directive. Even if there’s no such thing as zero risk, the combination of a shared security culture and advanced technological tools, particularly AI-based, makes it possible to build an effective defense against current and future threats.