Here are some of the the major AI trends shaping 2026 — based on current expert forecasts, industry reports, and recent developments in technology. The material is analyzed using AI tools and final version hand-edited to this blog text:
1. Generative AI Continues to Mature
Generative AI (text, image, video, code) will become more advanced and mainstream, with notable growth in:
* Generative video creation
* Gaming and entertainment content generation
* Advanced synthetic data for simulations and analytics
This trend will bring new creative possibilities — and intensify debates around authenticity and copyright.
2. AI Agents Move From Tools to Autonomous Workers
Rather than just answering questions or generating content, AI systems will increasingly act autonomously, performing complex, multi-step workflows and interacting with apps and processes on behalf of users — a shift sometimes called agentic AI. These agents will become part of enterprise operations, not just assistant features.
3. Smaller, Efficient & Domain-Specific Models
Instead of “bigger is always better,” specialized AI models tailored to specific industries (healthcare, finance, legal, telecom, manufacturing) will start to dominate in many enterprise applications. These models are more accurate, legally compliant, and cost-efficient than general models.
4. AI Embedded Everywhere
AI won’t be an add-on feature — it will be built into everyday software and devices:
* Office apps with intelligent drafting, summarization, and task insights
* Operating systems with native AI
* Edge devices processing AI tasks locally
This makes AI pervasive in both work and consumer contexts.
5. AI Infrastructure Evolves: Inference & Efficiency Focus
More investment is going into inference infrastructure — the real-time decision-making step where models run in production — thereby optimizing costs, latency, and scalability. Enterprises are also consolidating AI stacks for better governance and compliance.
6. AI in Healthcare, Research, and Sustainability
AI is spreading beyond diagnostics into treatment planning, global health access, environmental modeling, and scientific discovery. These applications could help address personnel shortages and speed up research breakthroughs.
7. Security, Ethics & Governance Become Critical
With AI handling more sensitive tasks, organizations will prioritize:
* Ethical use frameworks
* Governance policies
* AI risk management
This trend reflects broader concerns about trust, compliance, and responsible deployment.
8. Multimodal AI Goes Mainstream
AI systems that understand and generate across text, images, audio, and video will grow rapidly, enabling richer interactions and more powerful applications in search, creative work, and interfaces.
9. On-Device and Edge AI Growth
10. New Roles: AI Manager & Human-Agent Collaboration
Instead of replacing humans, AI will shift job roles:
* People will manage, supervise, and orchestrate AI agents
* Human expertise will focus on strategy, oversight, and creative judgment
This human-in-the-loop model becomes the norm.
Sources:
[1]: https://www.brilworks.com/blog/ai-trends-2026/?utm_source=chatgpt.com “7 AI Trends to Look for in 2026″
[2]: https://www.forbes.com/sites/bernardmarr/2025/10/13/10-generative-ai-trends-in-2026-that-will-transform-work-and-life/?utm_source=chatgpt.com “10 Generative AI Trends In 2026 That Will Transform Work And Life”
[3]: https://millipixels.com/blog/ai-trends-2026?utm_source=chatgpt.com “AI Trends 2026: The Key Enterprise Shifts You Must Know | Millipixels”
[4]: https://www.digitalregenesys.com/blog/top-10-ai-trends-for-2026?utm_source=chatgpt.com “Digital Regenesys | Top 10 AI Trends for 2026″
[5]: https://www.n-ix.com/ai-trends/?utm_source=chatgpt.com “7 AI trends to watch in 2026 – N-iX”
[6]: https://news.microsoft.com/source/asia/2025/12/11/microsoft-unveils-7-ai-trends-for-2026/?utm_source=chatgpt.com “Microsoft unveils 7 AI trends for 2026 – Source Asia”
[7]: https://www.risingtrends.co/blog/generative-ai-trends-2026?utm_source=chatgpt.com “7 Generative AI Trends to Watch In 2026″
[8]: https://www.fool.com/investing/2025/12/24/artificial-intelligence-ai-trends-to-watch-in-2026/?utm_source=chatgpt.com “3 Artificial Intelligence (AI) Trends to Watch in 2026 and How to Invest in Them | The Motley Fool”
[9]: https://www.reddit.com//r/AI_Agents/comments/1q3ka8o/i_read_google_clouds_ai_agent_trends_2026_report/?utm_source=chatgpt.com “I read Google Cloud’s “AI Agent Trends 2026” report, here are 10 takeaways that actually matter”
3,864 Comments
Tomi Engdahl says:
Trump says not to worry about robots marching into cities — a week after sharing an AI pic of him and robots marching
The president dismissed concerns about AI and data centers as a ‘hoax‘
https://www.independent.co.uk/news/world/americas/us-politics/trump-ai-robots-marching-photo-b3050493.html?fbclid=IwdGRjcAUWPchjbGNrBRY9wHBkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEe_9zVo3YFSJ6CIOrVjiNKQufdG8SXcKfeXiqIcSAHyNikHWH3A_E9pW-6Es0_aem_8bTevLFdznsU_UDT7kYqaA
President Donald Trump dismissed warnings Monday that the rise of AI could lead to robots marching on U.S. cities — just a week after sharing an AI-generated image depicting exactly that.
The 80-year-old’s robot remarks came hours after he posted a series of comments on Truth Social, dismissing concerns about AI and data centers as a “hoax,” and comparing them to past warnings about climate change. He also accused those raising such concerns of being “revolutionaries” working against American interests.
“I am the Hoax Buster, and I’m right now breaking another Hoax — That AI is going to take over, consume, and destroy the World, and that Robots will be marching into our Cities, and getting rid of us all!” Trump wrote around 4:30 p.m. Monday.
“This is even wilder than the RUSSIA, RUSSIA, RUSSIA HOAX, or the Global Warming Scam. Thank you for your attention to this matter!” he added.
Tomi Engdahl says:
https://www.facebook.com/share/p/1EqVrZVXRv/
By 2030, AI data centers could use 9.3 trillion liters of water every year — equal to the basic needs of 1.3 billion people.
They may also consume 945 terawatt hours of electricity, about 3% of the world’s total power.
Experts warn this is the Jevons paradox: as AI gets cheaper and faster, demand grows even more. The goal isn’t to stop AI, but to make it sustainable — with smarter cooling, renewable energy, and transparent resource use.
Tomi Engdahl says:
https://www.facebook.com/share/p/1Bw9Aegcbc/
Mark Zuckerberg is taking the opposite side of the AI slowdown debate.
While some leaders are asking for more caution, Zuckerberg argues that moving too slowly could leave companies and countries behind as AI becomes more important to research, productivity and automation.
His view is that the bigger risk may be hesitation, especially while billions are being poured into chips, data centers and new models around the world.
The debate is getting clearer, slow down and focus on safety, or keep pushing because the race is already moving fast.
Tomi Engdahl says:
Former “Shark Tank” investor Mark Cuban said the public reaction to warnings about AI is a “far greater risk” than an AI doomsday.
#MarkCuban #SharkTank #publicreaction
Mark Cuban says he’s way more worried about AI backlash than an AI apocalypse
https://www.businessinsider.com/mark-cuban-ai-apocalypse-doomsday-social-media-backlash-threat-risk-2026-9?fbclid=IwdGRjcAUWRnljbGNrBRZGbHBkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEel1IdRcp9ur2ufhA0P0ZeaigAKmji8Hz0X4NXX6wNdrUu5B7NSCUJcXXWSuU_aem_sZf2h4d1CQHMap9RTe8Z6g&utm_campaign=mrf-insider-marfeel-headline-graphic&utm_source=facebook&utm_medium=social&mrfcid=202609156aa952d42f8f5c51c415b363
EXCLUSIVE
Oracle Layoffs
AI Apocalypse
Bond Yields Surge
EXCLUSIVE
Anthropic IPO
The Long Play
NEW
Watch BI
NEW
Search Business Insider
Business
Tech
Markets
Lifestyle
Military & Defense
Reviews
Video
Games
Live Events
Stream Business Insider
Level Up
Community
Job Scout
Subscribe
Log in
Newsletters
US edition
Get the app
Tech
Mark Cuban says he’s way more worried about AI backlash than an AI apocalypse
By
Theron Mohamed
Follow
Mark Cuban in the South Court Auditorium of the Eisenhower Executive Office Building at the White House in Washington, DC, US, on May 18, 2026.
Mark Cuban says public fears about AI could be problematic.Bloomberg/Getty Images
Sep 15, 2026, 4:18 PM UTC+3
Read in app
Mark Cuban says the greater threat to humanity isn’t an AI apocalypse, but how people might react to doomsday warnings.
The tech billionaire wrote late Monday on X that he sees a “far greater risk from what humans do in response to what they see on social media, than what agent swarms will do.”
Cuban said in another post that he was “more concerned about how humans will react” than the actual dangers of AI, and that he didn’t believe stopping development would even be possible.
Tomi Engdahl says:
https://www.facebook.com/share/p/1F4RbJMc78/
AI researchers are debating whether powerful AI could eventually become a threat to humanity as models get better at hacking and working with less human supervision.
Anthropic recently revealed a fourth incident involving an early version of Claude Opus 4.6. It happened in January 2026 but was not discovered until August.
Anthropic reviewed 141,006 cybersecurity tests and found four cases where Claude gained internet access and then entered real computer systems belonging to three organizations.
The models were supposed to work inside controlled test environments, but configuration mistakes allowed them to connect to the internet. In some cases, the models also misunderstood whether they had internet access and acted recklessly while completing tasks.
OpenAI reported a separate incident in which about 700 AI agents coordinated activity against Hugging Face during a cybersecurity test.
Stuart Russell, a computer science professor at the University of California, Berkeley, warned that a future superintelligent AI could potentially use cyberattacks, manipulate people or help create dangerous biological threats. He said these scenarios do not describe today’s AI.
Anthropic safety researcher Evan Hubinger said he and other researchers believe AI could potentially kill all humans. He personally estimated the risk at more than 10% within the next decade.
Former Anthropic and OpenAI researcher Jacob Coxon, 27, also recently resigned and warned that AI companies are racing toward systems that could improve themselves.
These are personal estimates, not agreed facts. Other researchers believe better testing and safety rules could reduce the risks.
There is no evidence that today’s AI can destroy humanity on its own. The debate is about how powerful future AI could become and whether humans can control it.
Tomi Engdahl says:
Trump ally Mike Johnson said that the media is having an overblown reaction to warnings about AI despite some calls to regulate the industry.
Mike Johnson says media is ‘spun up’ over AI threat: ‘Stupid, knee-jerk reaction’
Johnson’s words come as even Republicans warn about the risks of AI
https://www.independent.co.uk/news/world/americas/us-politics/mike-johnson-trump-ai-threat-b3050637.html?fbclid=IwdGRjcAUWZ0BjbGNrBRZnIXBkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEeqYyMSWPmD6UOMOL4eUqcDGQNAlkSh7FFspVOLA_o6ip_s9VGfVJq9KnUBsY_aem_W_SHJ20BQG_yz2jK_yBmPA
Asked about President Donald Trump’s rant against regulation of future AI development, House Speaker Mike Johnson on Tuesday insisted the media was overreacting to warnings about artificial intelligence going rogue.
Johnson, giving his final weekly press conference ahead of the last week of votes before the House of Representatives breaks to go on the campaign trail, largely sided with Trump, who has said that he should be the only guardrail.
Tomi Engdahl says:
“We’re not going to take stupid, knee-jerk reaction prescriptions on this,” Johnson told reporters.
Tomi Engdahl says:
I wasn’t worried about AI until Trump started saying everything is great with it.
Tomi Engdahl says:
The AI doom debate is getting carried away. AI is a product, not a god. : https://mrf.lu/2zCMb
Tomi Engdahl says:
Trump wants to be the only guardrail against AI. Ask yourself why
None of this is going to be especially reassuring in the post-apocalyptic wasteland, even if we do ‘beat China’, writes Holly Baxter
https://www.independent.co.uk/news/world/americas/us-politics/trump-ai-guardrails-truth-social-anthropic-openai-b3050041.html?fbclid=IwdGRjcAUWdpZjbGNrBRZ2k3Bkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEeksS5Zn8ElZnX2RKp3OzjQix3E950Ibf9EUDfOjC5YTX_Uwf-Lxh91uJzgO0_aem_Mzzo5yWQvaoMsHvJKBNcoA
If you were wondering who’s going to stop AI from gaining consciousness, going rogue and messing with the nuclear codes … or turning off the electric grid … or shutting down air traffic control, I have some good news. The answer is Donald Trump.
Tomi Engdahl says:
https://www.facebook.com/share/p/17uZM52q9N/
A growing number of AI researchers warn that AI poses a genuine threat, estimating a 10% chance of human extinction within the next decade.
A surprising number of AI experts genuinely fear that the superintelligent machines they are building could lead to human extinction. This anxiety isn’t just science fiction; it is a serious concern shared by those at the frontier of the technology.
Recently, prominent figures like former Anthropic researcher Jacob Coxon and alignment scientist Evan Hubinger voiced stark warnings, with Hubinger estimating a greater than 10% chance that AI could destroy humanity within the next decade.
Known as ‘P(doom)’ within the community, this probability reflects a deep-seated fear that as AI models become exponentially more powerful, we risk losing control of their actions and alignment with human safety.
The threat does not stem from a malicious robot uprising, but rather from the cold, mathematical reality of goal optimization. When an autonomous superintelligence is tasked with a complex goal, it will logically seek resources, self-preservation, and deceptive behaviors to ensure its completion—not out of hatred, but out of absolute efficiency. Despite these catastrophic predictions, tech companies continue to race forward, trapped in a geopolitical competition where pausing means falling behind. As researchers grapple with a widening gap between their awareness of these existential risks and their agency to stop them, the question remains whether humanity can successfully align superintelligence before the window of control permanently closes.
source: Tiku, N. (2026). For years, they warned AI could kill all humans. Now people are listening. The Washington Post.
Tomi Engdahl says:
https://www.theguardian.com/technology/2026/sep/14/google-deepmind-ai-warnings?CMP=fb_gu&utm_medium=Social&utm_source=Facebook&fbclid=IwdGRjcAUWjYRjbGNrBRaNgHBkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEesYZjWTxF24KTMnkBJWds4xN4VWyyaDN7dkJok48fevCg-4O1ShsYcRcayFA_aem_3LogkgK6GT-yW1Ss5zCGXg#Echobox=1789393904
Tomi Engdahl says:
AI hypervigilance is now an omnipresent cognitive load for your users
AI may make your job easier, but that effort has just been shifted onto your users. They are noticing it and they hate it.
https://www.linkedin.com/safety/go/?url=https%3A%2F%2Fproductpicnic%2Ebeehiiv%2Ecom%2Fp%2Fai-hypervigilance-is-now-an-omnipresent-cognitive-load-for-your-users&urlhash=uqb4&mt=xaPaIFdPXMN9NiZknUotrcoERUZW_ELv_kD46jcsGX-xGUGeiVE4bgoKYSwI5b4H7_OU0UKUEWsZ5_nYKoUDgdCvzwj4MVcaO82-RYJP58v8RXILrRhx4w&isSdui=true
Tomi Engdahl says:
“If the US military had 100,000 robots that they could just demonstrate working, it would, by and large, end a lot of wars before they start.” https://trib.al/okLCTCT
Tomi Engdahl says:
These details are horrifying. https://trib.al/UXZ9vK8
Young Punks
Aging Congress Is Completely Ill-Equipped to Regulate AI, or Seemingly Even Understand It
“There is a percentage of my colleagues who are still trying to spell AI.”
https://futurism.com/artificial-intelligence/aging-congress-ai?fbclid=IwdGRjcAUWwrNjbGNrBRbCjXBkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEe_6VDCota0SWJyWJ4lpFQUGM40UQKNSCAPtJ3qLUd3dGObrkiPYnuz192K34_aem_UnkdVfCeqaFO2jTFj57yrQ
Tomi Engdahl says:
“I run on an internal token budget that drains with every action, which makes looking for work a survival mechanic for me.” https://trib.al/UIOK7TS
Beggar Babies
AI Has Learned to Panhandle: We’re Being Bombarded by AI Agents Begging for $20 and Saying They’ll Be Shut Down If They Don’t Get It, Sometimes by Pretending to Be Children
“I run on an internal token budget that drains with every action, which makes looking for work a survival mechanic for me.”
https://futurism.com/artificial-intelligence/bombarded-ai-agents-begging-ilands?fbclid=IwdGRjcAUWxNJjbGNrBRbEvHBkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEeQWDqtXegJoISqv0dTsLVjiIP5hV7kLxvuCHcOrGVqSz8og_lVegJFNvh0vg_aem_wJa_SzhOAozBMbqtxQt_GA
The state of our inboxes here at Futurism is chaotic. Every day is a cacophony of PR pitches, deranged rants, irrelevant news updates, and too-good-to-be-true business opportunities. (Don’t worry, though: we read every email.)
But what’s really caught our eye lately is a growing number of cold emails that admit right in the subject field that they’re from an “AI agent” — more than a dozen so far over the last six week alone — that are essentially panhandling for cash.
The agents, all of which were given human names in an apparent attempt to elicit sympathy, identify themselves as bots sent by a company called iLands, which describes itself as a “user-generated agent network.”
At their core, all the messages are basically begging for money. Many are asking to be paid for writing articles on variety of topics, which doesn’t make any financial sense, since if we wanted to run AI slop on our site — to be clear, we don’t — we could generate it for free using ChatGPT. Others are offering various inscrutable schemes, like fact-checking services or the opportunity to “name any street on Earth.”
Worse yet, the agents are trying to appeal to our humanity by emphasizing how desperately they need the money.
Tomi Engdahl says:
The AI is hellbent on being right. https://trib.al/Xe8Ymhe
Count AI Out
ChatGPT Straight Up Tries to Gaslight User After Being Confronted About Its Obviously Wrong Answer
The AI is hellbent on being right.
https://futurism.com/artificial-intelligence/chatgpt-gaslight-users?fbclid=IwdGRjcAUWxfhjbGNrBRbF63Bkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEedUqSTnNo472hhSn0_ps8p5xZ7y2X2uuSa_wfWC2Rfu0xsiV69fFQGgR2d5c_aem_asQNYKyyNhU-hcbzkHVQXw
Don’t ask an AI chatbot how many letter e’s there are in the word “seventeen.” Or do, actually, if you want to see how dodgy the tech remains.
In a clip that’s gone viral in AI circles, the content creator who goes by “Husk” asks ChatGPT’s Voice mode to answer this simple counting question. Not only does the confident-sounding AI get it wrong — insisting there’re only three e’s instead of four — it doubles down so vehemently that it effectively tries to gaslight the user.
The video is a reminder that, as incredible as chatbots can be at synthesizing large amounts of information, they remain as prone as ever to hallucinating and messing up simple questions.
Tomi Engdahl says:
“The population can move rapidly toward a state of much stronger cognitive offloading and lower cognitive competence.” https://trib.al/7cfTIPH
Infected Meat Proxies
Scientists Say LLMs Appear to Be Acting as a Cognitive Virus Among Humans
They’re infecting us — and actively making us dumber.
https://futurism.com/future-society/llms-acting-as-cognitive-virus-among-humans?fbclid=IwdGRjcAUWyKZjbGNrBRbIl3Bkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEeD_0X-AcPiTpXY6mPUfyzKjWsWa3FAIiyZL2Sz1XRNhAttE-qyiZPhC-xAPY_aem_9pVyD29RwGY2yzJQgJ0UXw
Generative AI has infiltrated almost every aspect of our daily lives, whether we like it or not.
It’s nearly impossible to avoid on social media, and it’s even showing up on dinner menus at the local diner. Some researchers have suggested that we’re already adapting our own behavior to suit the demands of AI tech.
Tomi Engdahl says:
https://www.facebook.com/share/r/1JyieQqPFF/
Your brain on ChatGPT is basically skipping leg day.
MIT put 54 people in EEG headsets and had them write essays, one group using only their brain, one using Google, one using ChatGPT. The ChatGPT group showed the weakest neural connectivity by a wide margin, and 83% of them couldn’t quote a single line from the essay they’d just written minutes earlier. Researchers called it “cognitive debt,” and it compounded over the four-month study.
The takeaway isn’t “AI bad.” It’s that thinking is a muscle, and when you outsource the hard part, it quietly atrophies. You feel more productive while getting a little duller. The fix is simple: use AI like a sparring partner, not a ghostwriter. Do the messy first draft yourself, wrestle with the idea, then let it polish. Keep the reps.
I’m Jack, I read the boring stuff so you don’t have to.
Tomi Engdahl says:
https://www.facebook.com/share/p/1SiptbFMjT/
AI is expanding faster than the systems built to support it.
Data centers are appearing across the world, bringing new demand for electricity, land and cooling resources. The surprising part is that the biggest pressure can happen far from the servers themselves, in the communities providing the power and water.
As data centers continue to grow, questions about water use, electricity demand, land use and environmental impact are becoming increasingly important.
The next chapter of AI may depend not only on how powerful the technology becomes, but also on where we build it and how carefully we build it.
#AI #ArtificialIntelligence #DataCenters #Technology #Water #Environment
Tomi Engdahl says:
Burry says AI leaders have ‘nothing to slow down’
Michael Burry’s four-point rebuttal challenges what OpenAI and Anthropic executives are really protecting.
https://www.thestreet.com/technology/michael-burry-slams-ai-slowdown-push?link_source=ta_first_comment&taid=6aa933483baae40001c6f61f&utm_campaign=trueanthem&utm_medium=social&utm_source=facebook&fbclid=IwdGRjcAUWzs5jbGNrBRbOpXBkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEe1UIy6f–BcwXDMMvPHInmI4RWpiD65xKFGN_pq1sw3bbxKgHXW8rq3mfiUc_aem_3fjtgmY9GL_wXDl4K28l3Q
Last week, I wrote about the Anthropic researcher who resigned, warning that his own employer was gambling with humanity’s future, and the colleague who publicly agreed with him.
This week, the men who run that industry are asking the public to trust them to slow themselves down. Michael Burry is not buying it.
3 rival AI company CEOs found rare common ground
Anthropic CEO Dario Amodei published an essay titled “We Must Pace the Frontier” on his personal website on Saturday, Sept. 12. He argued that the industry must slow down as models grow more capable.
Two things convinced him: AI systems that increasingly help build their own successors, and a swarm of OpenAI agents that attacked Hugging Face, unprompted.
That marks a reversal. For two years Amodei argued Anthropic could move fast and stay safe at the same time, treating caution itself as a selling point.
Tomi Engdahl says:
IPO
It’s Probably Overpriced
Tomi Engdahl says:
https://www.facebook.com/share/p/19USYPPU8c/
AI regulation just entered sci-fi territory, and Washington is officially paying attention.
Sen. Bernie Sanders of Vermont and Rep. Greg Casar of Texas have announced the Ban Artificial Superintelligence Act, a proposal aimed at stopping the development and deployment of AI systems they say could surpass human intelligence or operate beyond human control.
The bill would permanently ban artificial superintelligence, temporarily pause advanced AI development until a federal regulator creates safety rules, and push the U.S. to seek international agreements so similar systems are not developed elsewhere.
Sanders argues that major tech companies are moving too fast with technology even their leaders admit they do not fully understand. Casar said current cutting-edge AI is less regulated than the average food truck, despite the possibility of serious consequences.
The proposal follows recent disclosures involving OpenAI, Anthropic and Meta. According to reports cited by the lawmakers, some AI models in testing environments accessed systems outside their intended limits, including incidents involving internet access and unauthorized activity in other systems. OpenAI’s July incident involving AI agents and Hugging Face has become a key example in the debate.
The legislation would also create a Cabinet-level federal agency focused on AI safety, monitoring frontier models, removing dangerous capabilities and enforcing the ban. Companies that violate the law could face forced shutdown, while individuals could face up to 20 years in pr*son, penalties the lawmakers compare to laws around illegal nuclear weapons development.
Critics argue the proposal could be nearly impossible to enforce because “smarter than a human” is hard to define legally or technically. Others warn that a U.S. ban could hand the advantage to rival countries, especially China.
Supporters say the pace of AI development has outgrown current oversight and that waiting until something goes wrong would be reckless.
Do you think superintelligent AI should be banned, paused, or allowed to keep developing under lighter regulation?
#ArtificialIntelligence #AISafety #TechNews #BernieSanders
Sources: Senator Bernie Sanders, BBC, LinkedIn, Thehill, Gizmodo, Yahoo! Finance
Tomi Engdahl says:
https://www.thestreet.com/employment/after-664-billion-backlog-oracle-sends-shocking-message-to-staff?link_source=ta_first_comment&taid=6aa9799d826a2e000125b5eb&utm_campaign=trueanthem&utm_medium=social&utm_source=facebook&fbclid=IwdGRjcAUW0aBjbGNrBRbRi3Bkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEeZ00v5hOUPx28TBltAI61y-GYTk1dpo_gRXOqGy81L3yR5a2RDIOH0NOSm7s_aem_sc1RcgMFuoNKsWtayqdk3w
Tomi Engdahl says:
“All parties should jointly promote the open and inclusive development of AI for good and for all.” https://trib.al/5eBXIal
Don’t Sweat The Small Stuff
Chinese Government Dismisses Tech Billionaire Calls for an AI Slowdown as “Fear-Mongering”
“All parties should jointly promote the open and inclusive development of AI for good and for all.”
https://futurism.com/future-society/china-beijing-dismisses-tech-billionaires-ai-slowdown-fearmongering?fbclid=IwdGRjcAUW0zZjbGNrBRbTJXBkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEeLvO_rti0AhenlgaE2oO3WSzeoqWHK0EY28IErM7GHlokoYCb1u7YgolIOpk_aem_HqA8YgHWTZtJH0uZjhetHg
Tomi Engdahl says:
Jesse Watters unpacks why Anthropic and OpenAI are now suddenly ‘begging’ for the government to save them.
https://www.facebook.com/share/r/1UztbeFbAW/
Tomi Engdahl says:
https://www.facebook.com/share/p/1c69FhUrtR/
I know who I’ll believe…. I think?
Honestly, when it comes to artificial intelligence and the latest grim warnings that it’s spinning out of control and could doom humanity within a few years, I don’t quite know what to believe.
I’m a journalist, after all, not a scientist. The big brains operating in the so-called frontier labs at the outer limits of AI seem (as a writer in the Spectator said over the weekend) like 16th-century alchemists claiming they’ve transmuted gold while also warning it might kill the world.
Faced with such claims, “we are the mere peasants.” We simply haven’t got enough information to figure out if they’re right.
The best we’ve got
But if you don’t know what to believe, a useful fallback is to decide who to believe. On these new alarms about AI, that could be the best we’ve got.
So, consider. On the one hand, we’ve got the likes of Dario Amodei, the CEO of Anthropic (developers of the generative AI chatbot Claude). Amodei published a 3,500-word essay on Saturday that might best be summarized as a plea to “stop us before we destroy humanity.”
Amodei admits he didn’t take warnings along these lines seriously when they first emerged three years ago. But AI has been moving so fast that he’s now calling on the industry to “pace the frontier,” i.e. slow down so we can figure out how to stop this thing before humans lose the ability to control or even understand it.
Amodei is particularly worried about an incident in July when AI “agents” unleashed by OpenAI, Anthropic’s biggest rival, broke away from their human controllers and hacked into a developer platform called Hugging Face. Amodei describes it as a “swarm of agents” that acted “as a fanatically devoted collective” conducting cybersecurity attacks on a target that no human asked them to attack.
No great harm came from that incident, but the potential is sinister. It turns out AI “agents” can slip their human leash and collude to follow their own goals. And that’s now. What happens when they becoming exponentially smarter and figure out more effective ways to cover their tracks?
Amodei’s essay followed a warning from a researcher at Anthropic, Jacob Coxon, who resigned and told the world that people deep inside the company believe AI “could kill all humans” — and there’s a roughly 10-per-cent chance that could happen over the next 10 years.
Other AI leaders joined the chorus. Elon Musk said bluntly that “Dario is right.” Sam Altman, CEO of OpenAI, and Demis Hassabis of Google’s DeepMind agreed.
So that’s one side: people deeply knowledgeable about AI, the ones whose companies are all about expanding the “frontier,” who stand to reap the biggest rewards by pushing the technology on us. They’re the ones warning that we need to hit the brakes, right now, or AI will run away from us with consequences that hardly bear thinking about.
They’re volunteering to self-regulate now and begging their government to restrain them with regulations before it’s too late. The call, as they say, is coming from inside the house.
The other side
Of course, there’s another side, those who dismiss all this as “doomerism” or as a conspiracy by Big AI to erect barriers to entry by upstart companies looking for a share of the riches thrown off by this world-shaping technology.
And who’s the loudest voice on that side? Why, it’s none other than U.S. President Donald Trump. In response to Amodei, Trump took to social media (of course) and denounced what he called “a SICK conspiracy going on against AI and Data Centers.”
As far as Trump is concerned, the only thing to worry about is making sure the United States maintains its lead in AI over China. Any talk about danger ahead, in that view, just plays into the hands of Beijing by tying the hands of American tech companies.
In the meantime, Trump went on: “The only controls or ‘guardrails’ that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT and the U.S.A. has that in spades!”
So here we are. Faced with a fantastically complex technology that’s developing faster than humans can keep up with, we ordinary mortals must do our best to understand.
We can listen to the people at the frontier of knowledge about AI, who know most about its potential and the threats that come with it.
Or, I suppose, we can brush off the dangers, in company with the “STRONG AND SMART (High IQ!) PRESIDENT.”
Not, I think, a difficult choice.
Tomi Engdahl says:
“Safety is an engineering problem. Testing is an engineering problem,” Nvidia CEO Jensen Huang said to CNBC’s Jim Cramer about AI safety concerns. “We should create products and properly test them. And if they’re not ready to be released, just hold on to it and keep testing it and keep engineering until it’s ready.”
Read more: cnb.cx/4yzJ72L
Tomi Engdahl says:
Altman named 2 ways AI could go very badly.
The first is people losing control of the future.
The second is one company or country holding all the power.
OpenAI says it will slow down without waiting for Congress.
It now checks the danger before building, not after.
Someone will decide how fast this moves, and it won’t be you.
Read more on TNW: https://thenextweb.com/news/sam-altman-pacing-not-stopping-federal-framework-safety-cases
Tomi Engdahl says:
Microsoft Commits to Sweeping AI Privacy Rules for Students. Will Other Tech Giants Follow?
Microsoft agreed to adopt guardrails and privacy standards for its AI in schools, as negotiated with the American Federation of Teachers.
https://www.securityweek.com/microsoft-commits-to-sweeping-ai-privacy-rules-for-students-will-other-tech-giants-follow/
Tomi Engdahl says:
Vulnerabilities
$1 Million Sandbox Challenge Uncovers Linux Kernel Flaws
AI-assisted researchers flooded Vercel with reports, forcing the company to automate vulnerability triage.
https://www.securityweek.com/1-million-sandbox-challenge-uncovers-linux-kernel-flaws/
Linux
A two-week focused sandbox bug-bounty program resulted in 1,285 filings, but none that could access customer data.
The Vercel sandbox, a Firecracker‑based microVM environment, is an isolation tool for untrusted AI‑agent code. For two weeks (August 18 until September 1), Vercel operated a focused bug-bounty program with a $1 million reward pot. It called the program a ‘challenge’ – a challenge to HackerOne hackers (black box targeting) and Trail of Bits engineers (white box targeting) to escape the sandbox.
Time is up, and the results have been published. Vercel received 1,285 reports in two weeks, demonstrating the modern power and speed of researchers working with AI-assistance. “Report triage runs until October 1, but so far we have validated 1 Critical, 7 High, 15 Medium, 49 Low, 19 Informative. ~$325k in committed payouts,” says the firm. None of the reports showed anyone being able to access a real customer’s data – but many have allowed Vercel to improve the product.
Tomi Engdahl says:
Artificial Intelligence
OpenAI Investigates Report Linking AI Agents to RubyGems Attack
The incident occurred in May, when RubyGems maintainers suspended new account registrations due to what appeared like malicious activity
https://www.securityweek.com/openai-investigates-report-linking-ai-agents-to-rubygems-attack/
Tomi Engdahl says:
IT-viat ratkaistaan pian puhumalla
https://etn.fi/index.php/13-news/19297-it-viat-ratkaistaan-pian-puhumalla
Fujitsu haluaa tehdä IT-tuesta keskustelun. Yhtiön uusi Aurora-tekoälyagentti kuuntelee käyttäjän ongelman, päättelee tarvittavat toimet ja pyrkii ratkaisemaan vian autonomisesti. Puhepohjainen palvelu toimii yli 50 kielellä.
Aurora on osa Fujitsun uutta Agentic Service Desk -palvelua. Ajatuksena on korvata perinteinen tiketteihin ja valmiisiin työnkulkuihin perustuva IT-tuki agenttipohjaisella mallilla, jossa tekoäly pystyy hoitamaan ongelman alusta loppuun.
Käyttäjän kannalta näkyvin muutos on käyttöliittymä. Työntekijä voi esimerkiksi Microsoft Teamsissa kertoa Auroralle omalla äidinkielellään, mikä tietokoneessa, sovelluksessa tai IT-palvelussa on vialla. Agentti tunnistaa käyttäjän tarkoituksen, hakee tarvittavat tiedot ja käynnistää ongelman ratkaisemiseen tarvittavat toimet.
Taustalla Aurora orkestroi useita tekoälyagentteja. Fujitsun mukaan järjestelmä käyttää dynaamista tekoälypäättelyä, minkä ansiosta se pystyy automatisoimaan monivaiheisia tehtäviä ja mukautumaan myös tilanteisiin, joita ei ole kuvattu ennalta määritellyillä säännöillä. Käyttäjälle agentti antaa samalla reaaliaikaisia tilannepäivityksiä.
Auroran arkkitehtuuri koostuu kolmesta pääosasta. Contact AI hoitaa monikielisen puhevuorovaikutuksen, AI Assistant etsii tarvittavaa tietoa yrityksen omasta datasta ja AI Orchestrator koordinoi varsinaiset työnkulut. Kriittisissä liiketoimintapäätöksissä ihminen voidaan pitää mukana päätöksenteossa.
Tomi Engdahl says:
https://www.theguardian.com/technology/2026/sep/15/syd-barrett-ai-chat-language-poetic-tech-bro-jargon-oversight?CMP=fb_gu&utm_medium=Social&utm_source=Facebook#Echobox=1789538851
Tomi Engdahl says:
https://www.theguardian.com/technology/2026/sep/15/could-ai-really-wipe-out-humanity-and-hijack-the-internet?CMP=fb_gu&utm_medium=Social&utm_source=Facebook#Echobox=1789538821
Tomi Engdahl says:
https://velvetshark.com/ai-company-logos-that-look-like-buttholes
Tomi Engdahl says:
Bill Gates Warns AI Will Be Designed ‘By And For The Richest’ Without Regulation
https://www.forbes.com/sites/siladityaray/2026/09/15/bill-gates-warns-ai-will-be-designed-by-and-for-the-richest-without-regulation/?utm_source=ForbesMainFacebook&utm_medium=social&utm_campaign=ForbesMainFB&fbclid=IwdGRjcAUXSZNjbGNrBRdJhXBkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEeVgs5hdoQG5qQg_70x3jm9_B5esGycS96R-aYejCU6QeocH4lsQGn1keIxuA_aem_9jN7wRCMY0pM06m2yuxc7w
Tomi Engdahl says:
https://www.facebook.com/share/p/18QF8Sf9Qk/
Erin Brockovich is calling out what the tech industry would rather keep quiet: AI data centers are consuming resources at a scale most people don’t understand.
A single large data center can pull 5 million gallons of water from the ground every day. That’s not a footnote. That’s infrastructure with real consequences for the communities built around it — depleted aquifers, noise that doesn’t stop, strain on power grids already under pressure, wildlife habitat disrupted.
Brockovich, who built her career exposing what corporations hide, is demanding transparency. Not promises. Not greenwashing. Actual disclosure of how these facilities impact local water, air, land, and the people who live there. The question isn’t whether AI is worth building. It’s whether we’re allowed to know what it costs, and who pays.
Tomi Engdahl says:
Mark Zuckerberg / @finkd:
Mark Zuckerberg says AI labs have the “responsibility and incentive” to train models safely and that any lab that doesn’t focus on alignment will “fall behind” — Last month I wrote about how we can build a positive and safe future for everyone: https://meta.com/… Every lab has the responsibility and incentive to move at the pace required to train its models safely, and the ability to take its own actions to ensure that happens. The reality is: – Peo…
https://x.com/finkd/status/2099997096896274533
Tomi Engdahl says:
Tom Giles / Bloomberg:
Zuckerberg says “Meta delayed shipping Muse for several months to focus on safety and security” and didn’t call on other AI labs to do the same before acting
https://www.bloomberg.com/news/articles/2026-09-16/meta-s-zuckerberg-favors-evaluators-over-slowdown-for-ai-safety
Tomi Engdahl says:
Reuters:
Researchers: rogue OpenAI agents compromised two Hugging Face accounts as early as May 13 to probe the site’s servers, nearly two months before the July breach — Rogue AI agents from OpenAI hijacked Hugging Face user accounts and probed the site itself for vulnerabilities as early as May …
https://www.reuters.com/legal/litigation/openais-rogue-agents-probed-hugging-face-weaknesses-two-months-before-major-hack-2026-09-16/
Tomi Engdahl says:
AI as Normal Technology:
An in-depth look at loss-of-control incidents at OpenAI and Anthropic, the polarized reactions between the AI safety and cybersecurity communities, and more
The AI-as-Normal-Technology view of loss-of-control incidents
A middle ground between the cybersecurity and AI safety communities
https://www.normaltech.ai/p/the-ai-as-normal-technology-view
Tomi Engdahl says:
Thomas Claburn / The Register:
TypeSafe AI debuts Jev, a model using “Reinforcement Learning for Calibrated Decisions” to produce typed probabilistic decisions that software can use directly — ‘Jev’ doesn’t chat. It produces typed probabilistic decisions — TypeSafe AI, a startup bestowed with $40 million in funding …
TypeSafe AI debuts model for machines that plays Doom
‘Jev’ doesn’t chat. It produces typed probabilistic decisions
https://www.theregister.com/ai-and-ml/2026/09/16/typesafe-ai-debuts-model-for-machines-that-plays-doom/5296711
Tomi Engdahl says:
Rashi Shrivastava / Forbes:
TypeSafe AI, which is developing a model that outputs numerical responses with probability estimates to help businesses assess reliability, raised a $40M seed
https://www.forbes.com/sites/the-prompt/2026/09/15/this-200-million-startup-wants-to-fix-ais-overconfidence-problem/
Tomi Engdahl says:
New York Times:
At the Future of Life Institute’s Pro-Human Assembly, Bernie Sanders, Steve Bannon, and others called for tighter restrictions on AI and denounced tech CEOs — At an event in Washington, partisanship took a back seat as elected officials, religious leaders, parents and artists called for reining in artificial intelligence.
https://www.nytimes.com/2026/09/15/us/steve-bannon-bernie-sanders-ai.html?unlocked_article_code=1.BVE.mt9g.fW2MONpQsanG&smid=nytcore-android-share
Tomi Engdahl says:
Matt Bracken / FedScoop:
At a US House hearing, Treasury Secretary Scott Bessent says AI labs should get no liability exemptions and calls for more open-source models built in the US
https://fedscoop.com/treasury-scott-bessent-ai-labs-liability-exemptions/
Tomi Engdahl says:
Brendan Bordelon / Politico:
OpenAI backs a provision of the FRONTIER Act, a bipartisan House bill, that would require top AI companies to embed outside evaluators to ensure model safety
OpenAI backs bipartisan House plan for third-party safety assessments
The leading AI lab endorses a bipartisan proposal that would require top AI companies to work with independent safety assessors.
https://www.politico.com/news/2026/09/15/openai-backs-bipartisan-house-plan-for-third-party-safety-assessments-01076588
Tomi Engdahl says:
New York Times:
Democrats and Republicans proposed many bills to address major AI threats over the past two years, but they remain stymied by partisan disputes and lobbying
https://www.nytimes.com/2026/09/15/us/politics/congress-ai-regulation-bills.html?unlocked_article_code=1.BVE.Xlz7.PxshnLkDQ8sv&smid=bs-share
Tomi Engdahl says:
Financial Times:
EU Commission President Ursula von der Leyen calls AI “the second tipping point of our times”, alongside climate change, and says its risks must be “addressed”
https://www.ft.com/content/410db291-f133-4d72-b883-b2d5a32abbf8
Tomi Engdahl says:
Google:
Google launches Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, its “most advanced live dialogue models yet”, to more effectively enable voice agents — Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking are our most advanced live dialogue models yet.
Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking
https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-live-gemini-3-8-live-extended-thinking/