Tech Trends 2026
B · Useful, limited Deloitte's 17th annual Tech Trends report, widely read by CIOs; it combines Deloitte's US surveys of 500 and 302 technology leaders with interviews and secondary data, useful context but light on method.
Summary based on the full text·Original: about 2 hours to read
In short
Deloitte's Tech Trends 2026 (December 2025) argues that enterprises have moved from AI proofs of concept to scaling, and names five trends: physical AI and robotics, an "agentic reality check", the economics of inference infrastructure, rebuilding the tech organisation for AI, and AI in cyber defence. Its recurring message is that most firms automate existing processes instead of redesigning them, which is why agent deployments lag pilots.
Key numbers
- 11%surveyed organisations with agentic systems in production, against 38% piloting (Deloitte, 500 US tech leaders)
- 35%organisations with no agentic strategy at all; a further 42% still developing one
- 93% vs 7%split of AI investment between technology and people, per Deloitte research on allocation patterns
- 8% to 13%expected rise in the share of tech budgets allocated to AI over two years
- 2 millionprojected workplace humanoid robots by 2035
As reported by the source.
English·PDF·opens mkto.deloitte.com
What it reports
- Agents: 30% of organisations exploring, 38% piloting, 14% with solutions ready to deploy, 11% in production; 42% still building a strategy and 35% with none.
- Investment imbalance: 93% of AI spend on technology and 7% on people.
- Inference economics: token costs fell about 280-fold, yet some enterprises face monthly AI bills in the tens of millions; Deloitte suggests on-premises becomes worth considering when cloud costs reach 60% to 70% of equivalent hardware.
- Tech organisation: AI's share of tech budgets expected to rise from 8% to 13%; nearly 70% of tech leaders plan to grow teams because of gen AI; only 1% report no major operating model changes.
- Physical AI: projection of 2 million workplace humanoids by 2035; agents treated as a "silicon-based workforce" needing onboarding, performance tracking and FinOps.
- Security: risks across data, models, applications and infrastructure, plus AI-based red teaming and threat detection.
How they got it
Deloitte's 2025 Emerging Technology Trends in the Enterprise Survey (500 US technology leaders, June-July 2025), its 2025 Tech Spending Outlook (302 US IT and procurement leaders), other Deloitte surveys, executive interviews and published sources such as Gartner and the Stanford AI Index.
Limits
- Survey samples are US-only and self-reported; several figures come from Gartner predictions rather than data.
- The 93/7 investment split is attributed to "Deloitte research" without a method.
- Long, broad and partly promotional; customer quotes come from Deloitte's own CIO Journal content.
Why it matters for companies
The agent pipeline numbers (38% piloting, 11% in production) are a fair benchmark for where most firms are, and the report's diagnosis is the common one: redesign the process before adding agents. The people-versus-technology spend split is worth checking against your own AI budget, since change management, training and role design are where pilots stall. The inference-cost section is relevant for any firm whose AI usage is growing faster than unit prices fall.
Summary based on the full text. Read the original for the full argument and data.Open PDF on mkto.deloitte.com