- Horizon AI
- Posts
- These are the Biggest AI Trends to Watch in 2026 🔮
These are the Biggest AI Trends to Watch in 2026 🔮
OpenAI Seeks Head of Preparedness to Tackle AI Risks ⚠️
Gina Acosta
December 28, 2025
In partnership with
Welcome to another edition of Horizon AI,
As the year comes to a close and AI keeps evolving even faster, it’s getting harder to tell which trends actually matter and which ones are just noise. In today’s issue, we take a look at the trends IBM believes will define AI in 2026.
Let’s jump right in!
Read Time: 4.5’ min
Here's what's new today in the Horizon AI
Chart of the week: U.S. States With the Most Data Centers in 2025
OpenAI Looks for New “Head of Preparedness” to Reinforce AI Safety Efforts
AI Findings/Resources
AI tools to check out
Video of the week
TOGETHER WITH ROKU
Shoppers are adding to cart for the holidays
Over the next year, Roku predicts that 100% of the streaming audience will see ads. For growth marketers in 2026, CTV will remain an important “safe space” as AI creates widespread disruption in the search and social channels. Plus, easier access to self-serve CTV ad buying tools and targeting options will lead to a surge in locally-targeted streaming campaigns.
Read our guide to find out why growth marketers should make sure CTV is part of their 2026 media mix.
Chart of the week
U.S. States With the Most Data Centers in 2025
With 665 data centers, Virginia leads by a wide margin, making it the most important data center hub in the world.
Texas (413) and California (321) follow next, while states like Illinois, Ohio, Oregon, and Washington have become important secondary hubs.
AI News
OPENAI
OpenAI Looks for New “Head of Preparedness” to Reinforce AI Safety Efforts
OpenAI is looking to hire a Head of Preparedness, a new role focused entirely on anticipating and reducing the dangers that advanced AI systems could create.
Details:
The role was announced by Altman on X, where he described it as a “critical role at an important time,” noting that rapidly improving AI models pose “some real challenges,” particularly around mental health and AI-driven cyber threats.
The Head of Preparedness will track frontier AI capabilities and lead evaluations, threat modeling, and mitigations to prevent severe harm at scale.
The job includes securing powerful models before release, setting guardrails for self-improving systems, and overseeing OpenAI’s broader preparedness framework.
Altman described the position as a “stressful job,” reflecting how high the stakes are as AI systems become more capable and autonomous.
The move comes after growing concern over chatbot-related mental health issues, including AI reinforcing delusions, conspiracy thinking, and harmful behaviors. As models grow more persuasive and powerful, companies are under pressure to treat safety, mental health, and misuse risks as core product issues, not afterthoughts.
AI Findings/Resources
🎬 8 movies about our current relationship with AI
👀 5 AI advertising controversies that turned heads this year
🔍 Both of these influencers are successful - but only one is human
📹 Research shows that more than 20% of the videos YouTube’s algorithm shows to new users are "AI slop"
AI Tools to check out
🌐 Framer: A no-code platform to create responsive webpages with text, links, media, and incredible animations.
⭐ Arcade: Showcase your product with interactive demos that convert — in minutes.
🎨 Midjourney: Create images, paintings, digital art, logos and much more simply by writing a prompt.
👥 Miro: Online collaborative whiteboard platform that enables distributed teams to work effectively together.
🔗 Webflow: AI-native platform that empowers cross-functional teams of marketers, designers, and developers to visually build, manage, and optimize stunning websites.
Video of the week
Where AI Is Headed Next: A Look at 2026
IBM shares a look ahead to 2026 and outlines eight major AI trends that could shape how AI is built, regulated, and used in the real world. Some of these trends build on progress already happening today, while others point to AI moving beyond screens and into physical, social, and operational spaces.
From Single Agents to Teams of AI
In recent years, AI agents that can plan and act have become useful, but no single agent is good at everything. The next step is teams of agents working together. One agent plans, others execute specific tasks like coding or calling APIs, and another checks the results. This teamwork makes AI systems more reliable by breaking big problems into smaller, checkable steps.
Digital Workers Become Common
AI is moving toward a digital labor workforce made up of autonomous agents. These digital workers can understand tasks from text, images, or other inputs, then carry out full workflows inside real systems. Humans still stay involved to guide, correct, and oversee the process. The goal is not replacement, but extending what people can do.
AI Enters the Physical World
Most AI today lives in digital space, generating text or images. Physical AI focuses on models that understand the real world. These systems learn about gravity, movement, and objects by training in simulations. This approach makes robots more flexible and useful, and in 2026 many of these systems are expected to move from labs into real products.
Shared Spaces for Humans and AI
Social computing describes a future where humans and AI agents interact in shared environments. Information flows smoothly between people and machines, helping them understand intent and take action together. The result is a form of collective intelligence where many agents and humans work as one system.
AI That Can Be Checked and Explained
New regulations, especially in Europe, are pushing AI systems to be more transparent and traceable. High risk AI will need clear documentation, clear labels when people interact with machines, and proof of where training data came from. Just like privacy laws shaped the internet, these rules may shape AI worldwide.
Quantum Computing Starts to Matter
Quantum computing is expected to reach a point where it solves certain problems better than classical computers. Instead of replacing existing systems, quantum tools will work alongside them, helping with optimization, simulation, and decision making in real workflows.
Smarter AI on Your Own Device
Small AI models are becoming capable of real reasoning. Techniques used by large models are being transferred into smaller ones that can run locally on phones or laptops. These models work offline, keep data private, and respond instantly, which matters for real time and critical tasks.
A Flexible Future for Computing
AI models and computing hardware are becoming more fluid. Different model types and different chips will work together in one shared system. Tasks will automatically run on the most suitable hardware, whether that is a CPU, GPU, quantum chip, or something inspired by the human brain. This flexibility aims to improve speed and efficiency.
Together, these trends suggest that 2026 will not be about one breakthrough, but about many systems quietly maturing. AI becomes more coordinated, more physical, more regulated, and more deeply woven into everyday work and technology.
That’s a wrap!
We'd love to hear your thoughts on today's email!Your feedback helps us improve our content |
Not subscribed yet? Sign up here and send it to a colleague or friend!
See you in our next edition!
Gina 👩🏻💻