Generative AI is experiencing explosive growth, driven by major players like OpenAI and emerging competitors. According to a recent Cloudflare analysis, traffic to generative AI services has surged by 251% in one year, reflecting widespread global adoption. This surge is not limited to AI platforms alone—mobile operators are also feeling the impact, facing skyrocketing network traffic and heightened cybersecurity challenges. At the Mobile World Congress (MWC) 2025, industry experts emphasized the need to adapt telecom infrastructure to support this technological revolution. However, alongside this rapid expansion, billions of cyberattacks are now targeting AI platforms. Here’s a breakdown of the key trends shaping 2024 and 2025.
You want to know the impact of generative AI on mobile operators at MWC25? Get our Digest4Day.
The rise of generative AI services: ChatGPT leads, DeepSeek and Grok/xAI gain ground
Since its launch in late 2022, ChatGPT by OpenAI has maintained its dominance, now ranking among the top 50 websites worldwide (up from the 200th position at the end of 2023). While ChatGPT continues to grow, other competitors are gaining traction:
- DeepSeek, a Chinese chatbot launched in January 2025, quickly climbed to 3rd place before stabilizing at 6th.
- Grok/xAI, developed by X (formerly Twitter), saw its traffic surge following the release of Grok-3 in February 2025, now ranking between 9th and 10th place.
Other platforms are carving out niches with specialized features:
- Character.AI, enabling interactions with virtual personalities.
- Suno AI, a music generation tool that creates songs from text.
- GitHub Copilot and Codeium, leading AI-powered coding assistants.
These innovations are fueling the widespread appeal of generative AI, now embedded in various industries, from entertainment to software development.
Global adoption with distinct regional trends
Cloudflare’s DNS traffic analysis reveals significant regional differences in generative AI usage:
- ChatGPT.com: The U.S. leads with 17% of traffic, followed by Germany (7%), Brazil, Indonesia, and India (4% each).
- DeepSeek.com: Primarily accessed in the U.S. (14%), China (11%), and Germany (10%).
- Grok.com: Strong presence in Asia, with 20% of traffic from the U.S., 8% from Hong Kong, and 6% from Germany, Japan, and Vietnam.
- Poe.com, a popular AI platform in Asia, sees Hong Kong leading with 29% of its traffic, followed by the U.S. (13%), Japan (6%), and China (6%).
AI-powered development tools like GitHub Copilot and Codeium are most popular in the U.S. (22%), followed by Germany, India, and Japan (5% each).
Overall, the U.S. accounts for 23% of generative AI-related requests, followed by India (8%), Brazil (5%), Indonesia (4%), and the Philippines (4%). Europe remains a key player, while Egypt stands out as the highest-ranking African country (32nd place).
A surge in cyberattacks: 197 billion malicious requests blocked
While the growth of generative AI is remarkable, it has also become a prime target for cybercriminals, leading to record-breaking DDoS attacks (Distributed Denial of Service). Between February 2024 and March 2025, Cloudflare blocked 197 billion malicious requests, including 39 billion specifically targeting generative AI services.
Key attack milestones:
- June 2024: The month with the highest number of blocked malicious requests.
- February 23, 2024: Record-breaking DDoS attack with 3.7 billion requests blocked.
- November 13, 2024: Second-largest attack, with 1.5 billion blocked requests.
- January 2025: Multiple DDoS campaigns peaked on January 29, with over 1 billion requests blocked in a single day.
Primary attack targets:
- 80% of blocked requests targeted general-purpose AI chatbots like ChatGPT and Grok.
- AI image-generation platforms faced the highest proportion of blocked traffic, with over 50% of their total traffic identified as malicious.
These figures highlight the increasing vulnerability of AI services and the urgent need to strengthen cybersecurity defenses against ever-evolving threats.