【德勤】2025年技术媒体与电信TMT预测报告132页_4mb

TMT Predictions 2025: Bridging the Gaps Summary

Core Content

Deloitte's TMT Predictions 2025 highlight a pivotal year for the Technology, Media, and Telecommunications (TMT) sector, characterized by eight critical gaps that must be addressed to fully realize the potential of generative AI (gen AI) and other emerging technologies. The report also introduces two new sections containing 11 additional mini predictions, focusing on updates from previous reports and rising trends in the industry.

Main Topics and Predictions

1. The 2025 "Gap Year" for TMT

• Generative AI Infrastructure and Monetization Gap: Companies are investing heavily in gen AI infrastructure (e.g., chips, data centers), but returns are still low, with investment being 10x higher than revenue.
• Data Center Electricity and Sustainability Gap: Data centers require vast amounts of power, creating a challenge for sustainable energy solutions and grid capacity.
• Gen AI Gender Gap: Women's adoption of gen AI is accelerating, with Deloitte predicting parity in the U.S. by year-end 2025 and in many European countries within one to two years. However, a trust gap remains due to concerns about data security and AI bias.
• Deepfake Trust Gap: The rise of deepfake content is eroding consumer trust. Solutions include immutable labeling and real-time detection to ensure authenticity and mitigate risks.
• Studio Gen AI Usage Gap: While some studios are experimenting with gen AI, widespread adoption for creative workflows is slow. Studios are more likely to adopt gen AI for operations and distribution.
• Autonomous Gen AI Agent Gap: Agentic AI is expected to grow, with 25% of gen AI users launching pilots in 2025 and 50% by 2027. These agents can improve productivity and efficiency in various sectors.
• Streaming Video Gap: Consumers are reducing the number of streaming services they subscribe to, leading to a shift toward content aggregation and intermediary distribution.
• Cloud Spending Gap: The FinOps movement is gaining traction, with companies expected to save billions by optimizing cloud costs and linking spending to business value.

2. New Predictions and Trends

• On-Device Generative AI: Smartphones with on-device gen AI capabilities may see a 5% growth in shipments in 2025, up from 2024. This could enhance user experience with intelligent assistants and conversational interfaces.
• Wireless Telecom Consolidation: In markets where regulators allow, consolidation of wireless telecoms is expected to accelerate, especially in smaller markets, to create a more sustainable ecosystem.
• Fixed Wireless Access (FWA): Despite a slowdown in the U.S., global FWA adoption is expected to grow by 20% annually in 2025 and 2026, driven by enterprise demand and new markets.
• 5G Standalone (SA) Delay: Deployment of 5G SA networks is lagging, with Deloitte predicting only fewer than 20 additional networks will be upgraded in 2025.
• RISC-V and Geopolitical Risks: RISC-V chips are becoming a key alternative for gen AI, but U.S. export restrictions may limit their use in certain markets.
• Quantum Computing and Cybersecurity: With the rise of quantum threats, companies are expected to quadruple their investment in post-quantum cryptography by 2025.
• Chiplets and Moore’s Law: Chiplets are set to drive growth in advanced packaging, with revenue expected to more than double from $7 billion in 2021 to $16 billion in 2025.
• B/OSS Modernization: Telcos are moving toward cloud-based and SaaS solutions for their business and operations support systems (B/OSS), with the global market expected to reach $70 billion by 2025.
• Silicon Photonics: Used in data centers for gen AI, silicon photonics chips are expected to grow at a 25% CAGR to reach $1.25 billion in 2025.

Rising Trends

• Generative AI and Cybersecurity: While gen AI introduces new cyber threats, it also offers opportunities for defensive solutions, such as improved monitoring and risk management.
• Women and Generative AI: The adoption gap is closing, but trust remains a key barrier. Tech companies must address data security, reduce bias, and increase women's representation in AI development.
• Open RAN: Despite initial optimism, the transition to a multivendor ecosystem is slower than expected, with no new multivendor Open RAN networks likely to be deployed in 2025.
• Women's Sports Investment: The professionalization and commercialization of women's sports are driving increased investor interest, with rising valuations and new structures for funding.

Key Insights

• Generative AI is a central driver of TMT innovation, but challenges around infrastructure, trust, and adoption remain.
• Cloud optimization through FinOps is critical for cost control and value creation.
• Streaming video and telecom consolidation are reshaping how content is distributed and how networks are structured.
• Women's participation in gen AI and sports is growing, but trust and representation issues must be addressed to ensure long-term benefits.
• Emerging technologies like chiplets, silicon photonics, and RISC-V are playing a pivotal role in shaping the future of AI and telecommunications.

Conclusion

The 2025 TMT Predictions emphasize the need for collaboration, trust, and sustainability in addressing the gaps that hinder the full realization of gen AI and other innovations. By focusing on these areas, the industry can create a more inclusive, secure, and efficient future for businesses and consumers alike.