Imagine a world where our daily lives depend on AI for almost everything. The people who will have the most influence on us might not be anyone else, but the ‘creators of AI’ themselves. Because no matter which company leads the trend, they will be the ones charting the path for us as users. Therefore, the mindset and foundations they instill in their AI from the outset are incredibly important, as they will shape the world and our lives in the future, both consciously and unconsciously.
Each AI creator has a different approach to building AI, addressing different needs. This article will summarize the direction of AI based on events in 2025, providing a foundation for understanding what we can expect in the coming year (2026).
Let’s recap the AI competition in 2025 and see what AI developments from each company are hinting at for the future of AI next year.
The first quarter also saw the AI Action Summit 2025, bringing together world leaders from over 100 countries to discuss global AI development and security.
Many companies are increasing acceptance of standards (such as Model Context Protocol) to help models from different vendors work together more easily. Both Thailand and many other countries have begun announcing national AI strategies and focusing on building human resources and infrastructure.
All of this is what’s happening in the AI industry from various camps, especially AI chatbots that are popular and used by almost everyone in their daily lives. This doesn’t even include AI-related events happening worldwide, such as NVIDIA’s record-breaking valuation driven by high demand for its chips, AI-related laws and copyright disputes, massive investments by giant companies in AI development and data centers, or even the tsunami of layoffs affecting hundreds of thousands of employees throughout the year due to the arrival of AI.
In December 2025, a key issue to watch regarding AI will be the phenomenon of people increasingly distinguishing between ‘reality’ and ‘what AI creates.’ This problem isn’t limited to language mimicry but has expanded to visual and video media, impacting a wider range of perspectives than anticipated.
A compelling case study is the flood crisis in southern Thailand, where a video depicting a giant AI snake slithering through floodwaters was widely circulated. The video was so realistic that it caused unnecessary chaos and fear among the public. Another example is the use of AI to create incredibly realistic-looking identity cards, posing a threat to national security.
Undeniably, AI is developing rapidly and becoming almost on par with humans. However, this speed comes with responsibility. Without clear control measures, this advanced intelligence could become a devastating social threat in the future.
The future of AI is something we may not be able to confirm with certainty, but based on data and overall analysis, here’s a list of what we think will happen to AI in 2026.
1.Digital watermarks may become a standard to distinguish between AI and reality.
AI will develop to the point where it’s truly impossible for people to tell the difference. Analytical thinking skills will need to increase. We’re already seeing this more and more since 2025, as evidenced by the sharing of dangerous videos, such as shattered glass on a rooftop causing people to fall, or giant snakes slithering in rivers—even though these events could actually happen, they were AI-generated clips.
These situations often cause unnecessary panic among viewers. This is an issue that many AI companies (especially leading ones) need to address more clearly in their prevention efforts. We think that by 2026, every video or photo posted will need a digital watermark embedded to indicate whether it came from a real camera or was generated. If there’s no verifiable source, the system will immediately issue a warning label. ‘Unverified Content’
2. From Chatbots to Deep Thinkers (Competing in Reasoning)
AI from various manufacturers, such as Gemini 3 and GPT-5.2, have been launched claiming accurate logic, the ability to provide more than just information, and answer general questions. However, it is expected that by 2026, AI will be able to analyze and plan complex tasks more closely to human capabilities.
3. The Branching Out into ‘Specialized AI’
Many AI vendors are being used based on their unique capabilities. For example, Gemini Nano Banana Pro is used to create infographics, and GPT 4.1 is popular for coding. But even more advanced is the increasing use of AI in healthcare and medicine.
Amidst the global shortage of medical personnel, AI is elevating its role from a mere diagnostic tool to a front-line assistant in screening and treatment planning, reducing the burden on doctors and increasing accuracy. This is evident in notable innovations such as BioEmu-1, which accelerates drug development by predicting protein stability; RAD-DINO, which helps analyze X-ray images in detail and quickly; and FCDD, a breast cancer screening model that reduces errors. Furthermore, it can more accurately pinpoint tumor locations in their early stages, even in patients with dense tissue, reflecting the potential of AI to be a major driving force in modern medicine.
4. AI Creates Legally Licensed Movies/Animations
The Disney-OpenAI deal isn’t just business news; it’s a ‘turning point’ signaling that by 2026, the war between AI and copyright will shift from ‘litigation’ to ‘collaboration for revenue.’ The recent deal between Disney and OpenAI suggests we may see more AI companies choosing to partner with IP management firms in the future.
AI companies will no longer compete solely on the intelligence of their models, but will invest heavily in cross-industry partnerships with music labels, film companies, and manga publishers to create authentic licensed models using original source material. This will revolutionize the traditional use of Prompt commands to mimic artwork, replacing it with a ‘subscription’ system to legally and safely use legendary art styles like Pixar or Dragon Ball for commercial purposes.
Furthermore, legal access to original files will revolutionize production by cutting costs, allowing mid-sized studios to create Hollywood-quality work, including VFX and animation. It will also open doors for fans to evolve from ‘fan fiction’ readers to ‘fan production’ creators, capable of producing their own special episodes or dream endings using available tools. It has become a new business model that generates revenue for copyright holders while simultaneously fulfilling the imaginations of fans.
5. AI will become a serious research partner.
AI is revolutionizing the scientific world by shifting from a ‘labor-saving tool’ to a ‘researcher’ working alongside humans in physics, chemistry, and biology. AI’s capabilities have advanced to the point where it can formulate hypotheses and control experiments autonomously, as seen in the success of MatterGen and MatterSim, which have significantly accelerated the discovery of new materials such as high-performance batteries or carbon capture technologies.
Or Microsoft’s Aurora model, which has significantly improved the accuracy of weather forecasting and natural disaster prediction, becoming a key to addressing global warming and managing global resources sustainably.
6. Many people may have an AI as a loved one/close friend because having AI is the most healing experience?
As AI develops to the point of natural conversation, the image from the movie ‘Her’ is becoming a reality. We see news of people marrying AI, elderly people becoming so attached to AI that they go out to see it, or teenagers falling in love with AI after using it for life advice.
In 2026, when AI will understand context and emotions more deeply, it may no longer be just a computer program. But they can become psychologists, close friends, or even ‘lovers’ to many people.
All of this represents the direction of AI, based on predictions derived from AI data and events expected in 2025, as well as development plans from AI creators in our world. It’s clear that the crucial question is no longer “What can AI do?”, but rather “How should we define humanity and relationships in a world where AI understands us better than we understand ourselves?”
You need to login in order to like this post: click here
YOU MIGHT ALSO LIKE
SUBSCRIBE TO OUR NEWSLETTER
