Skip to main content
BTC / USDTCRYPTO107,400+2.19%ETH / USDTCRYPTO3,840+2.13%SOL / USDTCRYPTO182.40−1.99%BNB / USDTCRYPTO652.30+0.66%XRP / USDTCRYPTO2.2150+1.61%DOGE / USDTCRYPTO0.3850−1.79%TON / USDTCRYPTO5.240+2.34%AVAX / USDTCRYPTO42.60−2.07%LINK / USDTCRYPTO22.40+2.28%ADA / USDTCRYPTO1.0520−1.68%TRX / USDTCRYPTO0.3300+0.92%DOT / USDTCRYPTO8.420+2.93%BTC / USDTCRYPTO107,400+2.19%ETH / USDTCRYPTO3,840+2.13%SOL / USDTCRYPTO182.40−1.99%BNB / USDTCRYPTO652.30+0.66%XRP / USDTCRYPTO2.2150+1.61%DOGE / USDTCRYPTO0.3850−1.79%TON / USDTCRYPTO5.240+2.34%AVAX / USDTCRYPTO42.60−2.07%LINK / USDTCRYPTO22.40+2.28%ADA / USDTCRYPTO1.0520−1.68%TRX / USDTCRYPTO0.3300+0.92%DOT / USDTCRYPTO8.420+2.93%
Pricing
macroJun 26, 2026, 8:02 PM

AI Agent Market Set to Reach $300B by 2035, Memory Remains Challenge

New research projects the AI agent market will grow from $8 billion today to nearly $300 billion annually by 2035. However, a key limitation persists: AI agents struggle to remember information across sessions and understand data relationships.

According to recent projections from The Kobeissi Letter, the market for agentic AI is expected to expand dramatically, rising from approximately $8 billion currently to nearly $300 billion per year by 2035.

Despite rapid advancements in reasoning, coding, and writing capabilities, today's AI agents still face a fundamental bottleneck: memory. They have difficulty retaining information across sessions and comprehending connections between data points.

A new cohort of infrastructure firms is emerging specifically to address this challenge. The next frontier in AI development may shift from building smarter models to creating AI systems that can truly remember.

Source: The Kobeissi Letter