One year ago in Davos, executives panicked over DeepSeek’s surprise model launch. Investors questioned whether trillion-dollar AI bets were sustainable. The dominant question was whether the technology worked at all. 

That anxiety disappeared this year. As the World Economic Forum wrapped up today in Switzerland, the focus shifted to concrete products, timelines, and infrastructure costs. We tracked 10 interesting major AI announcements from the event.  

Here’s what was announced, who said it, and what these systems can actually do. 

1. OpenAI Confirms Hardware Device Launch Timeline 

OpenAI is “on track” to unveil its first consumer hardware device in the second half of 2026, according to Chief Global Affairs Officer Chris Lehane. Speaking at Axios House Davos on Monday, Lehane confirmed the device—developed with former Apple design chief Jony Ive—will be screenless. 

“The most likely window is later this year, depending on how things advance,” Lehane told attendees. The device marks OpenAI’s first move from software into hardware, positioning it to compete beyond chatbots to reimagine how people interact with AI in daily life. 

2. CEOs Put Dates on AGI Arrival 

Three of AI’s most powerful leaders offered competing timelines for when artificial general intelligence—systems matching human-level reasoning—might arrive. Google DeepMind CEO Demis Hassabis estimated 5-10 years, with a 50% probability by 2030. Anthropic CEO Dario Amodei maintained his 2026-2027 prediction, noting that models could soon handle elite-level reasoning in specialised fields. 

The diverging estimates reflect different definitions of AGI, but both agreed that the technology has moved from theoretical to imminent. The question is no longer “if” but “when.”

3. Meta Delivers Internal Advanced AI Models

Meta’s newly formed Superintelligence Labs delivered its first advanced models internally in January 2026, CTO Andrew Bosworth announced at Davos. The models—including a text-focused system codenamed “Avocado” and multimodal model “Mango”—are undergoing post-training for potential consumer release later this year. 

With 3.2 billion daily users across Instagram, Facebook, and WhatsApp, Meta has a distribution scale few competitors can match once these frontier models go public. 

4. Nvidia’s Infrastructure Warning: “Trillions of Dollars” Required

Nvidia CEO Jensen Huang called AI infrastructure the “largest infrastructure build-out in human history,” requiring “trillions of dollars” in spending across data centres, energy generation, and semiconductor manufacturing. 

“It’s wonderful that the jobs are related to trade craft—we’re going to have plumbers and electricians… all of these jobs, we’re seeing quite a significant boom and salaries have gone up. Nearly double,” Huang told attendees. 

5. Job Displacement Warnings from JPMorgan 

JPMorgan Chase CEO Jamie Dimon predicted AI will eliminate jobs within five years, telling CNBC that society may need to slow deployment. “We may have fewer jobs in five years,” Dimon said, adding that while new roles will emerge, the transition could cause significant disruption. 

Deloitte Global CEO Joe Ucuzoglu offered a measured counterpoint, noting AI will create new work categories even as it automates existing tasks. 

6. Anthropic CEO: Selling Chips to China is Like “Selling Nukes” 

Anthropic CEO Dario Amodei compared selling advanced AI chips to China to “selling nukes to North Korea,” urging the U.S. to maintain export restrictions, Bloomberg reported. Speaking at a closed session, Amodei described AI as a “superpower tool” that could fundamentally shift global power balance. 

The comments align with bipartisan concern in Washington about China’s AI capabilities. 

7. Isomorphic Labs Wants Clinical Trials by Year-End

Isomorphic Labs’ founder and CEO, Demis Hassabis, announced at the 2026 World Economic Forum that the company anticipates starting its initial clinical trials by the end of 2026. 

This timeline marks a bit of a backtrack from his earlier prediction at Davos 2025, where he’d pegged late last year as the moment AI-designed drugs would finally hit clinical trials. Even with the delay, the company is still doubling down on “digital biology”—essentially using its tech to map out biological systems and build new treatments from the ground up.

8. WEF Tracks Who Is Actually Using AI

The World Economic Forum released its “Proof over Promise: Insights on Real-World AI Adoption from 2025 MINDS Organizations” report, analysing real-world AI deployment across 30 countries and 20 industries. Developed with Accenture, the report showcased applications delivering measurable impact—from disease detection to energy grid optimisation. 

“AI offers extraordinary potential, yet many organisations remain unsure about how to realise it,” WEF Managing Director Stephan Mergenthaler said in the report. WEF also announced a new cohort of 20 organisations selected for the MINDS (Meaningful, Intelligent, Novel, Deployable Solutions) initiative. 

9. Nvidia’s Pitch: Europe’s "Once-in-a-generation" Shot at Robotics 

Jensen Huang is sounding the alarm for Europe, calling AI-powered robotics a “once-in-a-generation opportunity” that the continent can’t afford to miss. His advice to European leaders? “Get in early now.” Huang’s bet is that by fusing AI with Europe’s deep roots in physical manufacturing, the region could spark a massive resurgence in skilled trade jobs. 

This push comes at a critical time. While Europe has struggled to keep pace with the massive AI spending coming out of the U.S. and China, it still holds the upper hand in industrial automation and precision engineering. Huang is essentially telling them to play to their strengths before the window closes. 

10. Microsoft's Nadella Emphasises "Useful AI" to Build Public Trust 

Microsoft CEO Satya Nadella isn’t interested in AI for the sake of hype; he’s pushing industry leaders to focus on “useful” applications—especially in healthcare—to actually earn the public’s trust. During the panel, he noted that while AI’s impact could easily dwarf the invention of the internet, that potential doesn’t mean much if people don’t see tangible value in their daily lives. 

“We need to show up with real solutions to real problems,” Nadella told attendees, pointing to AI applications already reducing diagnostic errors and accelerating drug discovery.