AI in 2025: 25 significant events, such as huge agreements, concerns about bubbles, layoffs, talent battles, and more

ET Special: 25 Key Developments That Shaped AI in 2025 — And What 2026 May Bring

The year 2025 turned out to be a turning point for artificial intelligence, marked by massive investments, rapid innovation, job disruptions and growing concerns about bubbles and sustainability. Here are the key developments shaping AI today and what lies ahead.

? Investments, Funding & the AI Bubble

1. OpenAI hits $500 billion valuation

OpenAI became the first private AI company to reach a $500 billion valuation, turned fully for-profit, and loosened its ties with Microsoft. It is reportedly preparing for a $1 trillion IPO in 2026.

2. Circular funding raises bubble fears

Big tech firms like Nvidia, Microsoft, Oracle, AMD, SoftBank and OpenAI began investing in each other, creating circular funding loops. Nvidia briefly crossed a $5 trillion market cap, but volatility raised concerns about an AI bubble.

3. AI startups take over global VC funding

AI companies captured 53% of global venture capital in 2025, worth nearly $400 billion.
Startups are now hitting revenue milestones faster than ever, raising Series A rounds within months.

4. Youngest billionaires emerge from AI

Founders of AI startup Mercor became the world’s youngest self-made billionaires at age 22, highlighting how fast AI startups can scale.

5. Big Tech accelerates M&A

Google, Meta, Nvidia and others aggressively acquired or invested in AI startups.
Meta invested $14.3 billion in Scale AI, while Nvidia bought inference startup Groq for $20 billion.

? 2026 Outlook (Business & Valuations)

• Valuations likely to cool as focus shifts from hype to profits
• More mergers and acquisitions expected
• OpenAI and Anthropic may touch $1 trillion valuation
• Key question: will enterprise demand match massive data centre spending?

? Models, China & AI Innovation

6. The “DeepSeek moment”

China’s DeepSeek shocked the industry by training powerful models at very low cost, proving massive spending isn’t always necessary.

7. Rise of vibe coding & AI browsers

2025 saw growth in:

• AI coding tools
• Agent-based AI
• AI-powered browsers like OpenAI Atlas and Google Gemini Shopping

8. Chinese models catch up

Models like Qwen and Kimi are now performing close to Western AI models despite US chip restrictions.

9. Viral image tools explode

Tools like Nano Banana, Ghibli-style generators, and Sora went viral, raising copyright and privacy concerns.

10. Superintelligence debate heats up

OpenAI and Meta created teams focused on superintelligence. OpenAI said it could arrive within 10 years, though many researchers remain sceptical.

? 2026 Outlook (Models)

• AGI unlikely in 2026
• Sovereign AI models will grow
• Researchers will explore alternatives beyond transformers
• New “World AI” models are under development by Google, Fei-Fei Li, and others

?‍? Jobs, Layoffs & Talent Wars

11. Record tech layoffs

Over 122,000 jobs were cut in 2025.
Major layoffs included:

• Intel: 24,000
• Amazon: 14,000
• Microsoft: 9,000

12. Million-dollar salaries for AI talent

Top AI researchers received signing bonuses up to $100 million, even as layoffs continued elsewhere.

13. TCS layoffs mark historic moment

TCS laid off 12,000 employees (2%), its first major layoff ever, citing AI-led restructuring.

14. AI-driven hiring becomes mainstream

Around 75% of Indian firms now use AI tools for screening resumes and interviews.

15. Rise of hybrid workforce

Companies are now managing humans and AI agents together, marking the start of a hybrid workforce era.

? 2026 Outlook (Jobs)

• Hiring to remain cautious
• Non-AI roles likely to freeze
• 23% of jobs expected to change by 2027 due to AI and automation

?️ Infrastructure, Chips & Compute Wars

16. $400 billion yearly AI capex

Big tech firms spent $400 billion in 2025 on data centres and compute infrastructure.

17. Sustainability concerns grow

Global data centre power use may double by 2030. AI could consume up to 50% of that energy.

18. Google’s space data centre plan

Google announced Project Suncatcher, testing solar-powered satellites for AI workloads starting 2027.

19. Custom chips challenge Nvidia

Companies are building their own AI chips:

• Google TPUs
• Amazon Trainium
• Microsoft Maia
• Huawei Ascend

These threaten Nvidia’s dominance.

20. AI becomes cheaper

Token costs fell 90% since 2022, making AI more affordable. Experts say AI costs drop 10× every year, boosting usage.

? 2026 Outlook (Infrastructure)

• Hyperscaler capex may cross $600 billion
• 75% of spending will go to AI
• Nuclear, solar and satellite energy to power future data centres
• Google TPUs may slowly dent Nvidia’s lead

?? India’s AI Story

21. India launches ₹10,000 crore AI Mission

India rolled out subsidised AI compute at under ₹70 per hour, the cheapest globally, supported by 38,000+ GPUs.

22. Global AI firms expand in India

OpenAI, Anthropic, Figma and others are opening offices in India, drawn by its massive user base and talent pool.

23. $50 billion data centre push

Reliance, Adani, Google, Microsoft, Amazon and others will invest over $50 billion, expanding India’s capacity from 1 GW to 9–10 GW.

24. India becomes developer powerhouse

India now has 21.9 million developers and is expected to overtake the US by 2030 with over 57 million developers.

25. Rise of Indic AI models

Startups like Sarvam AI, Gnani AI, and others are building language models for Indian languages under the IndiaAI Mission.

? 2026 Outlook for India

• Shift from AI pilots to real-world deployment
• Push for sovereign AI will grow
• India likely to introduce a formal AI law
• Global spotlight on India during the India Impact Summit 2026 in New Delhi