NEW LAUNCHES

The latest features, products & partnerships in agents

• Salesforce to hire 1,000 sales execs for AI agent rollout

• AI giants are in a race to develop autonomous AI that controls your computer

• OpenAI’s ‘Operator’ tool could book flights and more by controlling your computer

• Microsoft’s Magentic-One is an AI agent that writes code and browses the web

• OpenAI to launch task-performing AI agents ‘Operator’ in January

• Startup Thesys aims to transform how users interact with AI agents

• Microsoft’s ‘Magentic-One’ framework directs multiple AI agents to complete your tasks

• DeepMind’s new AI agents framework mimics human-like reasoning

• Google confirms it did accidentally leak its AI agent Jarvis

IMPLEMENTATION

Announcements, strategies & tools

• How AI-API integration will enable competitive advantage for forward-thinking companies

• How to reinvent your company and maintain relevance in the age of agentic AI

• NVIDIA and Lenovo join forces in the name of ‘superhuman productivity’

• Productivity, security and the future of workplace dynamics

• Automation vs. agents — what’s the difference and which is best for your company?

• How AI can empower businesses to achieve ‘decision dominance’

• How AI-API integration will enable competitive advantage for forward-thinking companies

PREDICTIONS

Research, op-eds & key takeaways

• The most promising use cases for agentic AI

• Takeaways from the Forrester APAC Summit on agentic AI

• AI agents are the future tech companies have been promising

• AI21 CEO thinks unlocking AI agents means outgrowing the Transformer model

• How the rise of AI agents may signal an end to the app economy era

• AI agents will work independently by 2025, OpenAI predicts

• Microsoft AI chief: ‘ever present’ personal AI companions are coming

OPEN SOURCE

Launches, development, community & research

• What to know about OpenAI’s new Swarm AI agent framework

• How Dynamiq differentiates itself in the crowded AI agent landscape

• SAP expands open source AI support, turns Joule into collaborative agent

• New research explores how to train AI agents with an ‘evolving online curriculum’

IN OTHER NEWS

Compelling stories beyond the usual categories

• New Salesforce survey shows consumers increasingly receptivity to AI agents for service

• ‘Agent orchestration’ is the backbone of business ops in the AI era — here’s why

• Why ‘multi-agent AI’ is so much more powerful than LLMs

• AI shopping agents and the biggest trends shaping ecommerce in 2025

What’s happening in agents right now

Tech giants race to dominate the future of autonomous digital workers

Microsoft's early lead?

Microsoft has established a commanding position in enterprise AI agents, with their Copilot Studio platform now used by over 100,000 organizations. The platform's integration with 1,400 enterprise systems and access to 1,800 large language models through Azure gives Microsoft significant advantages in the emerging agent wars. Their latest development, Magentic-One, introduces an "Orchestrator" system that coordinates multiple specialized agents for general business tasks.

The company continues expanding Copilot's capabilities, recently introducing Copilot Actions for task automation and enhancing integration across Microsoft 365 applications. These updates include improved PowerPoint translation features, more sophisticated Excel analysis tools, and enhanced meeting management in Outlook.

OpenAI and Google join the race

OpenAI plans to counter Microsoft's enterprise dominance with "Operator," an AI agent capable of independently controlling computers and performing complex tasks. Set for release in January 2024, Operator represents OpenAI's push beyond basic text and image processing into autonomous system control.

Google isn't standing idle. The company is developing its own agent technology, joining Anthropic and other major tech players in the rush to build autonomous AI systems. Hardware manufacturers are also preparing for this shift - mobile device makers like HONOR are implementing agent capabilities in smartphones, while chip manufacturers develop specialized hardware support for AI agents.

The sales paradox

An interesting contradiction has emerged in the AI agent landscape. Salesforce, while promoting AgentForce as their AI automation solution, is hiring 1,000 new sales executives and product personnel to support it. This highlights a peculiar dynamic where companies pushing workforce automation are simultaneously expanding their human workforce to do so.

Startups target the interface challenge

While tech giants battle over broad agent capabilities, startups are tackling other aspects of the agent ecosystem. Thesys, a California startup, recently secured $4 million in seed funding to develop AI-driven user interfaces. Their platform aims to bridge the gap between advanced AI capabilities and user-friendly experiences, with 300 businesses already on their waitlist.

Consumer trust and adoption

Recent Salesforce survey data reveals complex consumer attitudes toward AI agents. While 30% would work with an AI agent for faster service, trust remains a significant concern. About 72% of consumers report trusting companies less than a year ago, and 73% want explicit disclosure when communicating with AI agents.

These trust issues haven't deterred enterprise adoption. Major organizations like Aflac and NASA are already implementing agentic AI solutions. In software development, where talent shortages persist, Gartner predicts AI agents will write most code within three years.

Technical reality check

Despite rapid advancement, current AI agent systems achieve only about 50% of human accuracy levels. Microsoft's Magentic-One system, while promising, still faces significant limitations in reliability and consistency. The technology requires careful implementation and robust error handling to be practical for business use.

Hardware integration presents another challenge. While mobile manufacturers and chip makers are building infrastructure to support AI agents, standardization and optimization remain works in progress. The industry must balance the push for autonomous capabilities with practical limitations of current technology.

The underground movement

An intriguing trend has emerged: one in five workers are now "underground" AI users, adopting agent technology without official sanction. This grassroots movement suggests significant pent-up demand for AI assistance, even as organizations work to develop formal implementation strategies.

The race to dominate AI agents marks a new phase in tech industry competition. While Microsoft's enterprise integration gives them an early advantage, OpenAI's upcoming Operator release and Google's development efforts could shift the balance. Meanwhile, companies like Salesforce and startups like Thesys are carving out their own niches in this rapidly evolving landscape.

The next year will likely determine which approaches to AI agents gain traction in the enterprise. With major players investing heavily and startups targeting specific challenges, the technology is moving from experimental to practical deployment. The key question isn't whether AI agents will become mainstream, but which companies will shape how they're implemented and used.

We publish daily research, playbooks, and deep industry data breakdowns. Learn More Here

Agents on the podcast

Our latest conversations around AI agents