Interested in AI? Check out our other AI whitepapers from this series:
Table of contents
Ayming Institute :the think tank of the Ayming Group.
The Ayming Institute (AI) aims to help leaders in the private and public sector gain a deeper understanding of the evolving global economy by focusing on three areas.
The first area is sustainability. We believe that the environment and social responsibility are critical issues for businesses today. For this reason, our content aims to help companies integrate these issues into the way they make decisions.
The second area is business development. Through our content, we wish to help companies to develop a stronger business culture and a sustainable approach to growth.
The third area is the people side of the business. With our content, we want to support individuals as they navigate their careers, learn new skills, and find ways to contribute in a world that is constantly changing.
Our strongest commitment is to help organizations better understand how markets are changing, and how they can build better businesses as a result. We aim to do this by providing analysis of the global economy’s transformation; sharing our insights through thought-provoking publications, and engaging business leaders in conversations about the economic changes that are affecting all of us.
AI and the Cyber Arms Race
Cybercrime is one of the world’s major growth industries. For years, the speed, scale, and sophistication of cyberattacks have been steadily increasing. With advances in AI, the threat is accelerating and amplified. But, as in other industries and applications, GenAI is a double-edged sword. It can be wielded by defenders or attackers, as a potent weapon in the cybersecurity arms race.
The digital downside
As our transactions, services and human interactions gravitated online, crime followed. The global pandemic and shift to hybrid/home working provided further impetus to this significant downside of the digital revolution.
In France, for example, the spread of teleworking is cited as a major factor in a 400% increase in cyberattacks between 2020 and 2023.1 Many organizations worldwide have rapidly shifted applications and data to cloud-based services, creating new vulnerabilities. Since 2020, 79% of companies with cloud data had experienced at least one cloud breach.2 Similarly, an IBM review found 82% of violations involved data stored in the cloud.3 Data loss and leakage, data confidentiality and accidental exposure of credentials were prime concerns.
The digital supply chain is another major risk – highlighted in 2021 by the log4j vulnerability in Java-based utilities (such as those generating error 404 ‘not found’ messages). Hackers had an opportunity to take control of countless servers worldwide. They were quick to respond: a third of the scanning activity to exploit the new weakness occurred within 30 days of release.
A 2023 investigation of data breaches also found almost a quarter involved ransomware – usually organized crime groups encrypting data and demanding a ransom to return or unlock it.4 Although external actors were responsible for most breaches, 19% involved employees causing unintentional or intentional harm. Even AI-enhanced systems need safeguarding against human ignorance and indifference: 74% of all breaches had a human element – through error, misuse of privileges, stolen credentials, or social engineering (criminals duping victims into disclosing details, clicking on links or attachments, or visiting websites that install malware).
All sectors of industry, and public bodies, are targets. Big corporations offer the largest attack surface and spoils, but they devote more resources to protection. SMEs tend to be more vulnerable and lack cybersecurity professionals. However, skills shortages are widespread. A worldwide workforce survey in 2022 estimated there were 4.7 million people in those roles, but a shortfall of 3.4 million.5
The enemy does not lack human or capital resources. An analysis of Microsoft’s vast data pool showed a 51% increase in cybersecurity alert signals to 65 trillion over the 12 months to June 2023.6 That’s over 750 million per second. The most common targets were passwords. Phishing emails accounted for a quarter of attacks.
The tech giant tracked more than 300 separate actors, including 160 backed by nation states and 50 ransomware groups.7 China, Russia, North Korea, and Iran launched cyberattacks against organisations in 120 countries.
Counting the cost
If cybercrime was concentrated in a single country, its economy would be the third largest, after the US and China. The damage in 2021 was estimated to be $6 trillion – double the 2015 toll.8 Cybersecurity Ventures expects this to grow 15% a year to $10.5 trillion by 2025 – more than the global trade in illegal drugs.
Global ransomware – the fastest-growing cybercrime – is estimated to have cost $20 billion in 2021, a 57-fold increase since 2015. By 2031, that could exceed $265 billion.
According to IBM, the average cost of a data breach in 2023 was $4.45 million, a 15% increase in three years.9 Costs vary significantly with the size of organization (as well sector, country, and other factors). Studies from 2021 suggest the average cyber claim was
€775,000 for mid-sized companies in France, but 6% faced costs of €3-10 million.10 In the US meanwhile, small businesses suffered over 700,000 attacks in 2020 and $2.8 billion in damages.11
The tab for cybercrime is long, and includes lost data, stolen money, theft of intellectual property, business disruption and interruption, forensic investigations, restoration of data and systems, and reputational harm. Lawsuits may follow with clients or suppliers. Future sales can suffer.
As the threat – and target – increase, organizations must invest more in defence. The global cybersecurity market was worth $3.5 billion in 2004. Spending over the five years to 2025 is projected to hit $1.75 trillion.12 By then, total global data storage could exceed 200 zettabytes. Everywhere from smartphones to private and public IT infrastructures, and IoT (Internet of Things) devices, half that data could be in the cloud.
Aiding the enemy
Since GenAI’s release into the mainstream at the end of 2022, the digital and business worlds have experienced one of their largest disruptions. Security and risk management leaders cannot afford to ignore it. Through 2025, securing GenAI alone could require a 15% spike in spending on application and data security.13 But many businesses are embracing the latest tech without managing associated risks. Risk analysts reported in 2023 that 93% of companies recognized the data privacy and other risks associated with using generative AI inside the enterprise, but only 9% felt prepared to manage the threat.14
By 2024, 85% of cybersecurity professionals already attributed increased attacks to GenAI use by bad actors.15 A third or more believe GenAI will amplify the volume and velocity of attacks (33%), make phishing undetectable (37%), and further compromise privacy protection (39%).
In the hands of criminals and spies, new-generation AI tools will be better at finding weaknesses and exploiting them at scale and speed.16 Not only can the enemy operate with greater efficiency, attacks are likely to become more sophisticated.
Phishing emails have already become more plausible (fewer spelling and grammatical errors, better designs, more personalized). AI enables more refined methods for assessing and selecting targets, scanning broad networks, user groups and social media for ammunition and vulnerabilities, analysing behaviours, evaluating the effectiveness of attacks, and adapting them.17
GenAI makes it easier to rewrite a known attack code to avoid detection. It was a hot topic on the dark web during 2023. Hackers boast of using ChatGPT – in one case recreating malware strains from research publications, including a Python-based stealer that can search and retrieve specified file types across a system.18
GenAI can also churn out a barrage of deepfakes, sham content, and feigned credentials to target human credulity – the weakest link in cybersecurity. Criminals have already replicated the voices of family or friends in messages and phone calls. Other dangers include virus- like malware that self-evolves, and polymorphic code undetectable by existing security measures.
The case for defence
The same generative capabilities could bolster cyber defences – for example, by creating data to simulate attack patterns never seen before,19 acting as a co-pilot for security teams (or software developers),20 and creating scripts and virtual scenarios to identify network anomalies and vulnerabilities.21
GenAI’s greatest potential – and application so far – lies in threat identification. Analysts can spot attacks faster, and better assess their scale and likely impacts as it filters alerts and rejects false positives.22 These tools should become more dynamic and automated. Many experts expect AI algorithms to provide real-time threat analysis if not increasingly autonomous responses.23
Leading tech companies and cybersecurity specialists, along with other research institutions, are working on smart solutions. Using AI to analyse trillions of security alert signals, Microsoft believes its GenAI Security Copilot can help close skills gaps. In a trial, novice analysts increased the accuracy of their responses by 44% and were 26% faster across all tasks.24
GenAI’s large language models arm security teams with the power to sift vast amounts of data for insights. But the intricacies of cybersecurity datasets may be limiting. A predicted transition to more agile, small-scale models could yield more tailored and actionable intelligence.25 Real-time data training might prove a ‘secret weapon’ delivering an even swifter response to ever-changing threats.
IP on the line
In addition to the many costs of cybercrime, there is another potentially far-reaching economic impact on innovation and investment. Intellectual property (IP) theft can stunt or destroy a business or fledgling industry, as well as skewing competition between companies or countries.
The extent of the problem is hard to measure. In 2018, it was estimated IP theft represented at least a quarter of the then $600 billion total annual cost of cybercrime.26 Analog-style smuggling of manila folders soon gave way to proven cyber tactics in over 80% of cases – from malware to phishing.
Again, the trend is upward. In the US, enforcement actions over IP theft were up 21% in 2023. The total cost of American IP theft increased 36% to more than $1.1 billion.27 Most publicly reported cases involve employees or ex-employees selling secrets to competitors or foreign states, but project partners and rival companies also instigate thefts.28
Nation states are behind most ‘advanced persistent threats’ (APTs). Military defence and security related secrets are not the only targets. The 2021 Microsoft Exchange breach, which compromised 300,000 organisations worldwide, was blamed on China.29 Its state espionage provides foreign corporate secrets to domestic businesses – the source of half of all Chinese-owned technology, according to a 2018 estimate.
A US annual review of global IP protection and enforcement cited other states not already in the cyber-hostile Top 4. Trade partners of primary concern included Argentina, Chile, India, Indonesia and Venezuela. Meanwhile, Algeria, Canada, Egypt, Mexico and Vietnam needed to improve their IP laws and enforcement. 30
Multinationals may be big targets, but so are start-ups and SMEs at the leading edge of technology. A common pitfall (familiar to Ayming) is to under-estimate the innovative value of technical know-how and progress.31 Businesses risk not only missing out on financial support, such as tax credits for R&D, but also failing to protect their intellectual assets. Another common mistake is not to document or record ideation and progressive steps en route to patentable breakthroughs.
Proactive security and R&D
GenAI poses threats and opportunities across the cybersecurity landscape. By streamlining training and detection, its applications have a vital role to play in combating AI-enabled attackers. Moreover, the technology can make cybersecurity more proactive.
Although many cybersecurity companies are exploiting GenAI in other areas, full automation of containment, eradication and recovery operations is unlikely in the medium term.
Cybersecurity R&D is and will continue to be vital in the AI-accelerated arms race to stem the colossal haemorrhaging of cash and IP to criminals and hostile states. The availability of national and EU subsidies reflects the sector’s priority status. Grant schemes and R&D tax credits have supported the development of a variety of technologies by Ayming clients, such as French start-up Qevlar32 and award-winning Sekoia.33
Growing risk awareness and regulation around data security are spurring cybersecurity literacy in the board room and investment in tools, teams and specialist services. Major organisations recognise the need to optimize their resilience and their cybersecurity operations. But many SMEs need to follow suit: 83% are unprepared for the financial after- damage and 91% of small businesses do not have cyber professional liability insurance.34
Just as cybersecurity demands a strategy and budget, adequate resources need to be dedicated to IP protection. This is essential not only to secure intellectual assets, but also investors’ confidence – and their contributions to the cost of protection. Without it, a company may lose its value and even collapse. Putting off action until a project is finished is dangerous. A young company must ask at the outset, and regularly thereafter, what ideas and work add value, are attractive to investors, and merit protection.
How to protect IP?
Ayming research shows that security risks to IP are widespread and challenging to manage. In our 2024 Innovation Barometer international survey, 85% of respondents had been embroiled in conflicts over their IP in the last five years.
With a total estimated cost of €500 billion, these disputes ranged from large-scale industrial espionage to partnership breakdown. Findings included:
- 38% had shared confidential research among partners without
- 34% saw their own innovations patented by competitors;
- 29% had their new products copied by others; any protection, such as a non-disclosure agreement (NDA);
- 29% said their ideas had been disclosed;
- 38% had incurred significant IP litigation costs (rising to 44% among larger companies).
Apart from sound cybersecurity protocols and NDAs, patents are the main means of protecting company secrets. However, today’s rapid innovation cycles can overtake the patenting process, which consumes time and resources that growing companies may lack. Often, they are unable to mark each technological advance. Yet, proving precisely when the company achieved such innovation can be crucial in a dispute with a rival.
New technology provides a solution. Ayming IP Secure is a collaborative Web 3.0 platform that creates a tamper-proof, time-dated record of each contribution to an innovative project. Based on blockchain, it strengthens protection by providing legal evidence of anteriority in a few clicks. Filing can be integrated with everyday business tools, saving time, and even automated.
In contrast with other methods, Ayming IP Secure is low-cost, allows unlimited number and size of submissions, in all formats, and is valid internationally.
Authors
Julien Portejoie
Market Leader Software Publishers, Ayming France
Trained as a computer engineer, Julien has been supporting software publishers and platforms in their financing and innovation strategies for over 12 years. With feedback and significant benchmarking in this market, he also supports Ayming’s consulting and sales teams to provide them with new knowledge, whether technical or methodological to deliver targeted client solutions.
Christopher Casta
Market Leader ESN/ICT, Ayming France
Christopher is a Market Leader at Ayming France. Trained as an engineer in IT, telecoms and electronics with a doctorate in image processing, Christopher has supported companies in the digital field for more than 10 years with their R&D financing and innovation strategies.
1 https://www.linkedin.com/pulse/cybercriminalité-chiffres-clés-à-connaitre-en-2023-cyber-cover-fr
2 https://www.ninjaone.com/fr/blog/7-statistiques-sur-la-cybersecurite-que-chaque-pme-et-msp-doit-connaitre/
3 https://www.ibm.com/fr-fr/reports/data-breach
4 https://www.verizon.com/business/resources/reports/dbir/
5 https://media.isc2.org/-/media/Project/ISC2/Main/Media/documents/research/ISC2-Cybersecurity-Workforce-Study-2022.pdf
6 https://dcmag.fr/le-monde-entre-dans-une-nouvelle-ere-de-cybermenaces/
7 https://www.microsoft.com/en-us/security/security-insider/microsoft-digital-defense-report-2023
8 https://cybersecurityventures.com/cybercrime-damage-costs-10-trillion-by-2025/
9 https://www.ibm.com/fr-fr/reports/data-breach
10 https://www.linkedin.com/pulse/cybercriminalité-chiffres-clés-à-connaitre-en-2023-cyber-cover-fr/?originalSubdomain=fr
11 https://www.strongdm.com/blog/small-business-cyber-security-statistics
12 https://cybersecurityventures.com/top-5-cybersecurity-facts-figures-predictions-and-statistics-for-2021-to-2025/
13 https://www.gartner.com/doc/reprints?id=1-2G6U1RDP&ct=240109&st=sb
14 https://www.securitymagazine.com/articles/100028-data-privacy-among-top-concerns-for-workplace-generative-ai-use
15 https://www.cobalt.io/blog/cybersecurity-statistics-2024
16 https://www.code-intelligence.com/blog/cybersecurity-trends
17 https://www.technology.org/2024/02/13/global-costs-of-cyberattacks-in-2024/
18 https://www.bain.com/insights/generative-ai-and-cybersecurity-strengthening-both-defenses-and-threats-tech-report-2023/
19 https://blogs.nvidia.com/blog/generative-ai-cybersecurity/
20 https://www.crowdstrike.com/cybersecurity-101/secops/generative-ai/
21 https://www.sangfor.com/blog/cybersecurity/what-is-generative-ai-cybersecurity
22 https://www.bain.com/insights/generative-ai-and-cybersecurity-strengthening-both-defenses-and-threats-tech-report-2023/
23 https://www.code-intelligence.com/blog/cybersecurity-trends
24 https://www.microsoft.com/en-us/security/blog/2023/11/15/microsoft-unveils-expansion-of-ai-for-security-and-security-for-ai-at-microsoft-ignite/
25 https://www.weforum.org/agenda/2024/02/what-does-2024-have-in-store-for-the-world-of-cybersecurity/
26 https://www.wipo.int/sme/en/top-10-intellectual-property-mistakes-smes-entrepreneurs.html
27 https://www.cyberhaven.com/guides/top-ip-theft-statistics
28 https://www.digitalguardian.com/blog/ip-theft-definition-and-examples
29 https://www.rusi.org/explore-our-research/publications/commentary/dual-approach-tackling-ip-theft-and-ip-loss
30 https://www.cyberhaven.com/guides/top-ip-theft-statistics
31 https://www.wipo.int/sme/en/top-10-intellectual-property-mistakes-smes-entrepreneurs.html
32 https://techfundingnews.com/qevlar-ai-secures-e4-5m-to-use-gen-ai-to-detect-cyber-threats/
33 https://www.lemondeinformatique.fr/actualites/lire-sekoiaio-elu-startup-it-2023-du-monde-informatique-92185.html
34 https://www.ninjaone.com/fr/blog/7-statistiques-sur-la-cybersecurite-que-chaque-pme-et-msp-doit-connaitre/
