Report: AI Expands Corporate 'Shadow IT'
Contrary to expectations that AI would consolidate software stacks, it is accelerating SaaS sprawl and the use of unmanaged "shadow IT," according to a new report from Torii. The 2026 benchmark report found that 61% of applications in the enterprise are unmanaged. This trend increases governance and security risks for large organizations.
- The procurement cycle for new enterprise software, particularly AI, often spans three to six months, creating a significant barrier to rapid innovation and adoption. Enterprise buyers are increasingly forming buying groups for AI solutions, moving away from individual persona-based sales strategies. - Chief Revenue Officers (CROs) are leveraging AI to gain a competitive edge by identifying emerging market opportunities and risks proactively. However, a Gartner survey shows only 7% of sales organizations achieve 90% or higher forecast accuracy, a gap AI is expected to close by analyzing historical deal patterns and customer behavior. - Agentic AI architectures represent a shift from single-task AI to multi-agent systems that can collaborate to achieve complex goals. These systems are orchestrated to decompose large tasks, assign specialized roles to different agents, and manage their communication and coordination to handle complex, real-world workflows. - The Bay Area remains the epicenter of AI funding, with over $200 billion raised since 2020; however, the landscape has shifted, with corporate investors like Amazon, Google, and Microsoft now accounting for 40% of total AI funding. Early-stage funding has contracted while mega-rounds for companies with proven revenue, like OpenAI and Anthropic, have increased, raising the bar for Series A startups to need $5M+ in annual recurring revenue. - The primary obstacles to enterprise AI adoption are poor data quality and fragmented data systems, with many procurement leaders citing messy, inconsistent data spread across multiple ERPs and spreadsheets as a major barrier. This can cripple AI algorithms, which require clean, centralized data to function effectively. - As startups scale, founders must transition from hands-on operators to strategic leaders who empower their teams. This involves shifting from personal problem-solving to building systems and processes that can function independently, a necessary evolution as a company grows beyond 30 employees. - Effective founders often use time-blocking and "No Extra Time" (NET) techniques to maximize productivity. This involves scheduling all activities, including deep work and personal commitments, on a calendar and pairing tasks like listening to podcasts during a commute or taking meetings during a walk. - In the current venture climate, investors are prioritizing capital efficiency, with competitive Series A rounds in the Bay Area requiring startups to show year-over-year growth of at least 50% and a burn multiple below 2.0. Specialized accelerators like HF0 and AI Grant are gaining prominence for AI founders due to their focus on compute resources and technical mentorship.
