Enterprises demand orchestration for AI agents

- The shift to orchestrated AI agents is driven by the need to manage the "coordination overhead" that arises when multiple specialized agents must interact. This complexity grows with the number of agents, creating challenges in communication, resource sharing, and task dependency that can lead to system deadlocks or inefficiencies. Platforms like LangGraph and CrewAI are emerging to model these complex agent workflows, managing state and inter-agent communication. - Governance frameworks for AI agents are evolving to focus on runtime controls, continuous monitoring, and clear audit trails to manage the risks of autonomous actions. Key components of these frameworks include defining an agent's scope and authority, establishing human-in-the-loop oversight for high-risk decisions, and ensuring that agents inherit permissions based on the principle of least privilege. This is a significant shift from traditional AI governance, which primarily centered on the risks of model outputs rather than the actions taken by the model. - In the adtech space, the deprecation of third-party cookies is forcing a move towards using first-party data and privacy-preserving technologies like Google's Privacy Sandbox. This initiative uses techniques like federated learning and differential privacy to enable targeted advertising without tracking individual users across websites. For advertisers, this means a greater focus on contextual targeting and new measurement models for attribution. - A notable trend in programmatic advertising for 2026 is the move toward platform ownership, with agencies and media buyers opting for white-label DSPs to gain more control over margins, data, and technology dependencies. This is part of a broader "Custom AdTech Renaissance" where organizations are building more of their own infrastructure to increase transparency and reduce reliance on third-party platforms. - For engineering leaders scaling their organizations, a key strategy is to structure teams into small, autonomous "pods" or "squads" with clear ownership over specific features or products. This approach, advocated by leaders from companies like Gong and Harness, aims to maintain velocity and individual accountability as the team grows. It's also critical to scale the architecture in parallel with the team to avoid bottlenecks. - The UK tech startup ecosystem saw a rebound in venture capital investment in 2025, reaching $23.6 billion, a 35% increase from 2024 and the first annual growth in four years. AI startups were a major driver, attracting a record $7.9 billion. However, funding in early 2026 has seen a 19.35% drop compared to the same period in 2025, with seed and early-stage funding contracting more significantly than late-stage investments. - In Formula 1, teams are preparing for a major regulatory overhaul in 2026 with the introduction of new chassis and engine rules. The new power units will feature a significant increase in electrical energy, with drivers having to adapt to new driving styles, including aggressive lift and coast techniques to manage energy recovery. - Recent local news in London includes a notable increase in the region's unemployment rate to a five-year high following job losses in January. Additionally, there are ongoing concerns at London's Nestle ice cream plant regarding a potential sale and the scrapping of a U.S. tariff.