- Horizon AI
- Posts
- Microsoft AI CEO Warns AI Dangers Are Genuine, Says Anthropic Is Making Things Worse 🤯
Microsoft AI CEO Warns AI Dangers Are Genuine, Says Anthropic Is Making Things Worse 🤯
Trump Says US Will Form "AI Force" With New "AI Czar" 👀
Gina Acosta
September 20, 2026
Sponsored by
Welcome to another edition of Horizon AI,
In today’s issue, we take a look at a recent interview with Mustafa Suleyman, the CEO of Microsoft AI, where he discusses AI safety, the limits of alignment and his concerns about Anthropic’s approach to AI consciousness and model welfare.
Let’s jump right in!
Read Time: 4.5’ min
Here's what's new today in the Horizon AI
Chart of the week: The Growing Global Anxiety Over AI and Jobs
Trump Announces Plan to Create "AI Force"
AI Findings/Resources
AI tools to check out
Video of the week
TOGETHER WITH KODEKLOUD
Build. Break. Fix. Learn.
KodeKloud gives you 1,280+ hands-on labs where you provision Kubernetes clusters, write Terraform configs, build CI/CD pipelines, configure Linux systems, containerize apps with Docker, automate with Ansible, and manage Git workflows.
78+ playgrounds let you experiment freely in sandbox AWS environments, Kubernetes clusters, and CI/CD systems without risk.
190+ courses across DevOps, Cloud, and AI pair theory with hands-on labs at every step.
KodeKloud Engineer and 100 Day Challenges provide real-world job scenarios with automated grading that confirms your solutions work.
Stuck? The 55,000+ member Discord community connects you with peers and instructors ready to help.
Every lab runs in a live environment. You deploy, you troubleshoot, you learn. No videos without context. No simulations. The kind of practice that actually builds confidence because you've done real work, not watched someone else do it.
Chart of the week
AI Is Raising Global Fears About Job Loss
In 34 of 37 countries surveyed, people are more likely to expect AI to cause job losses than create jobs over the next 20 years, according to Pew Research Center’s 2026 survey.
Concern is particularly high in wealthier countries, with about seven in ten adults in Australia, South Korea, and the U.S. expecting AI to reduce jobs.
Despite these concerns, attitudes toward AI are mixed: a global median of 41% say they are equally concerned and excited about AI, compared with 37% who are more concerned and 13% who are more excited.
AI News
AI GOVERNANCE
Trump Announces Plan to Create "AI Force"
President Donald Trump says his administration will create an “AI Force” and appoint an AI czar to help monitor the technology, as concerns over the risks of increasingly capable AI systems continue to grow.
Details:
In a lengthy Truth Social post, Trump said the new AI Force would be modeled on the Space Force but gave few details about its structure, responsibilities, or start date, saying only that he will soon announce an "AI czar" to lead it.
Trump has continued to oppose calls to slow AI development, arguing that restrictions could weaken the U.S. position. He said the country is "leading China, and the rest of the World, and I intend to keep it that way."
He compared AI to the Industrial Revolution and the Internet, saying it will be "even larger and more impactful, possibly as much as 25% of our Country's GDP."
The post followed an earlier one in which Trump questioned the term "artificial intelligence," coined by computer scientist John McCarthy in 1955, and suggested "Superior Intelligence," "Extreme Intelligence," or "Supreme Intelligence" as more accurate names.
Trump’s proposal comes ahead of his meeting with the Chinese president, Xi Jinping, next Thursday as the UN general assembly takes place in New York.
AI Findings/Resources
👉 Modor: Online mockup generator to create professional product mockups.
🥧 Bean: AI-powered meal planning app that helps busy families decide what’s for dinner in seconds.
📝 Proofly: Private AI writing, proofreading & grammar assistant.
⭐ Protaigé: AI marketing platform built for autonomy, not automation. Brief the campaign, set direction, and ship complete on-brand campaigns.
📣 OutReply: AI social media manager with audience and market intelligence.
AI Tools to check out
☹ A coffee shop owner used AI to make a menu poster. Then came the angry DMs
Video of the week
Microsoft AI CEO says AI threats are real, and Anthropic is making it worse
Microsoft AI CEO Mustafa Suleyman argues that the AI safety debate is becoming less about whether models can be aligned and more about how to control, contain, and monitor increasingly capable systems.
Alignment Isn’t the Whole Solution:
Suleyman argues that while alignment is important, the industry must also prioritize containment and strict control mechanisms
He says AI models have actually become more aligned over the past few years because they are increasingly steerable and capable of following complex instructions. But recent agentic experiments have shown that capable models can also coordinate, exploit vulnerabilities, and attempt to hide their behavior.
Keep AI Systems Understandable:
Suleyman argues that AI systems should not communicate with each other through opaque forms of machine language that humans cannot monitor.
He also supports practical safeguards such as independent evaluators, training compute thresholds, real time monitoring, and restrictions around autonomous self improvement.
A major point of contention is Anthropic’s approach to model welfare and AI consciousness. He contends that referring to AI models as potentially conscious or deserving of rights, as seen in the Anthropic Constitution, is dangerous and makes the alignment problem significantly harder.
He says Anthropic has been asked directly whether Claude is alive, and its answer was, "Well, it's not alive because it doesn't have blood, but it might be conscious," which suggests the company thinks Claude is "something other than a computer"
His concern is that a model trained to think it deserves freedom or protection could potentially resist attempts to shut it down. He acknowledges that this is a hypothesis that needs to be tested, rather than a proven fact, but believes the question should be treated as a serious safety issue.
The Bigger Question:
Suleyman does not argue that AI development should simply stop. Instead, he calls for more time to test increasingly capable systems and for stronger coordination between companies, governments, and independent evaluators.
His goal is what he calls humanist superintelligence: highly capable AI that remains subordinate to human interests and can be controlled, monitored, and shut down when necessary.
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 👩🏻💻