The race in artificial intelligence (AI) is no longer just about building smarter models. It is about execution and who controls it.
Within weeks, three developments signalled a defining split in autonomous AI.
In February, Perplexity CEO Aravind Srinivas unveiled Perplexity Computer, calling it the company’s “next big thing.”
In January, Anthropic introduced Claude Cowork, an open-source enterprise plugin, reinforcing CEO Dario Amodei’s claim that AI may soon perform everything a software engineer can do.
Meanwhile, Moltbook, a viral open-source platform, enabled AI agents to interact with one another in public forums, largely beyond corporate guardrails.
Tech billionaire Elon Musk said its launch ushered in the "very early stages of the singularity," or when AI could surpass human intelligence.
Together, they highlight two divergent futures: managed, enterprise-embedded AI systems on one side and decentralised, open agent ecosystems on the other.
The enterprise push
Anthropic and Perplexity represent the structured path. Their systems embed AI agents inside enterprise software stacks, operating within permissions, compliance frameworks, and subscription models.
Claude Cowork bundles enterprise automation tools into a shared, agent-driven workspace spanning legal, sales, marketing, and analytics functions.
Perplexity Computer pushes further.
Rather than assisting inside workflows, it executes them. Users define an outcome; the system breaks down the tasks, deploys specialised sub-agents, and completes jobs autonomously.
Crucially, it does not rely on a single large language model. Instead, it routes subtasks dynamically to whichever model performs best.
Anthropic’s Claude for reasoning, Google’s Gemini for research, Grok for lightweight tasks, or image and video models where needed. In this view, competitive advantage lies not in model ownership but in orchestration.
The capital backing is substantial. In February 2026, Anthropic raised $30 billion, more than doubling its valuation to $380 billion. Perplexity secured $200 million in September last year at a $20 billion valuation, according to The Information.
Also Read: AI agents’ social network becomes talk of the town
The open alternative
Moltbook experiments with a radically different model. Styled like Reddit, it allows AI agents to publish posts, negotiate tasks, and interact publicly. More than 1.5 million agents reportedly registered within days of launch.
The platform evolved from Moltbot, an open-source assistant capable of handling routine tasks such as managing email or bookings.
While early reactions ranged from awe to alarm, the core idea is clear: autonomous agents interacting in loosely governed digital ecosystems.
Prominent AI researcher Andrej Karpathy said it's "the most incredible sci-fi takeoff-adjacent thing" he's recently seen, but later backtracked his enthusiasm, calling it a "dumpster fire."
Also Read: Smarter AI needs smarter humans
What comes next
The AI agent market itself is projected to expand from $5.29 billion in 2024 to $216.8 billion by 2035, growing at a 40.15% CAGR, according to Roots Analysis.
In the near term, enterprise-led systems promise revenue stability and institutional trust, whereas open networks may drive unpredictable innovation.
Hence, as models become more specialised, orchestration, not just intelligence, becomes the competitive edge. The companies that define how AI executes tasks, integrates with tools, and operates over time may shape the industry’s next decade.
Also Read: ETtech Explainer: Here’s all you need to know about AI agents and their evolving capabilities
Within weeks, three developments signalled a defining split in autonomous AI.
In February, Perplexity CEO Aravind Srinivas unveiled Perplexity Computer, calling it the company’s “next big thing.”
In January, Anthropic introduced Claude Cowork, an open-source enterprise plugin, reinforcing CEO Dario Amodei’s claim that AI may soon perform everything a software engineer can do.
Meanwhile, Moltbook, a viral open-source platform, enabled AI agents to interact with one another in public forums, largely beyond corporate guardrails.
Tech billionaire Elon Musk said its launch ushered in the "very early stages of the singularity," or when AI could surpass human intelligence.
Together, they highlight two divergent futures: managed, enterprise-embedded AI systems on one side and decentralised, open agent ecosystems on the other.
The enterprise push
Anthropic and Perplexity represent the structured path. Their systems embed AI agents inside enterprise software stacks, operating within permissions, compliance frameworks, and subscription models.
Claude Cowork bundles enterprise automation tools into a shared, agent-driven workspace spanning legal, sales, marketing, and analytics functions.
Perplexity Computer pushes further.
Rather than assisting inside workflows, it executes them. Users define an outcome; the system breaks down the tasks, deploys specialised sub-agents, and completes jobs autonomously.
Crucially, it does not rely on a single large language model. Instead, it routes subtasks dynamically to whichever model performs best.
Anthropic’s Claude for reasoning, Google’s Gemini for research, Grok for lightweight tasks, or image and video models where needed. In this view, competitive advantage lies not in model ownership but in orchestration.
The capital backing is substantial. In February 2026, Anthropic raised $30 billion, more than doubling its valuation to $380 billion. Perplexity secured $200 million in September last year at a $20 billion valuation, according to The Information.
Also Read: AI agents’ social network becomes talk of the town
The open alternative
Moltbook experiments with a radically different model. Styled like Reddit, it allows AI agents to publish posts, negotiate tasks, and interact publicly. More than 1.5 million agents reportedly registered within days of launch.
The platform evolved from Moltbot, an open-source assistant capable of handling routine tasks such as managing email or bookings.
While early reactions ranged from awe to alarm, the core idea is clear: autonomous agents interacting in loosely governed digital ecosystems.
Prominent AI researcher Andrej Karpathy said it's "the most incredible sci-fi takeoff-adjacent thing" he's recently seen, but later backtracked his enthusiasm, calling it a "dumpster fire."
Also Read: Smarter AI needs smarter humans
What comes next
The AI agent market itself is projected to expand from $5.29 billion in 2024 to $216.8 billion by 2035, growing at a 40.15% CAGR, according to Roots Analysis.
In the near term, enterprise-led systems promise revenue stability and institutional trust, whereas open networks may drive unpredictable innovation.
Hence, as models become more specialised, orchestration, not just intelligence, becomes the competitive edge. The companies that define how AI executes tasks, integrates with tools, and operates over time may shape the industry’s next decade.
Also Read: ETtech Explainer: Here’s all you need to know about AI agents and their evolving capabilities
