testified.ai Logo

Google Forms Strike Team to Chase Anthropic's AI Lead

The competition in the AI industry has reached a new level of intensity with the formation of a Google DeepMind strike team, personally led by co-founder Sergey Brin. The team's explicit goal is to catch up to Anthropic's advanced coding capabilities, a move that validates recent user polls showing a massive shift toward Claude. This development is part of a larger landscape of strategic investments, security concerns, and major infrastructure projects shaping the future of AI.

Google Rallies DeepMind to Out-Code Anthropic

According to a report from The Information, Google co-founder Sergey Brin is spearheading a new Google DeepMind strike team to improve Gemini's code-writing abilities. This initiative is a direct response to the growing consensus, reportedly shared by researchers inside DeepMind, that Anthropic's Claude models have a significant lead in agentic coding.

Brin has framed this effort as the most direct path to creating self-improving AI systems, a long-standing goal in the field. The new group is reportedly led by research engineer Sebastian Borgeaud. This news coincides with reports that the NSA is using an internal-only Anthropic model, further highlighting the perceived capabilities of its technology.

Anthropic's Momentum Solidified by Investment and User Adoption

Anthropic's strong position was reinforced by two major events. First, Amazon expanded its collaboration, committing up to an additional $25 billion for a total of $33 billion invested. This deal secures up to 5 gigawatts of compute capacity for training and deploying Claude models on AWS.

Second, a reader poll from The Neuron newsletter showed Claude overtaking ChatGPT as the most-used AI tool, with a nearly 2-to-1 margin among respondents. Users cited superior coding quality, better long-form writing, and even alignment with the company's ethical stance as reasons for their preference.

The poll results suggest a significant portion of the power-user market is shifting allegiance, with tool choice increasingly becoming a reflection of both features and brand values.

Funding, Infrastructure, and Corporate Maneuvers

The broader industry continues to see massive capital flows and strategic shifts. Here are the key highlights:

  • Funding Rounds: Jeff Bezos' AI startup, code-named Project Prometheus, is reportedly close to finalizing a $10 billion funding round. Meanwhile, Factory AI, a coding agent company, is now valued at $1.5 billion after a $150 million raise, and Recursive Superintelligence, a four-month-old startup, raised $500M at a $4B valuation.
  • OpenAI's Stargate: A $500 billion infrastructure project involving OpenAI, Oracle, and SoftBank is actively developing seven sites across the US. The planned capacity is over 9 gigawatts, enough to power the equivalent of 20 million Nvidia H100 GPUs.
  • OpenAI Departures: Three leaders have recently left OpenAI: Kevin Weil (CPO), Bill Peebles (Sora co-creator), and Srinivas Narayanan (CTO for B2B Applications).

Market Challenges: Security, Supply Chains, and Billing

Despite the rapid growth, the industry faces several pressing challenges.

Security and Data Privacy

Vercel suffered a data breach originating from a compromised employee account on another AI product. The company stated that affected customers have been contacted. In other news, Atlassian announced it will enable default data collection to train AI on its 300,000 customers, a move that has sparked concern among users.

n8n (Workflow Automation) Logo
n8nTestified Badge
4.7/5

Supply Chain and Platform Costs

An ongoing memory chip shortage, fueled by the data center boom, is now projected to last through 2027. This is expected to drive up prices for electronics like smartphones and laptops. In response to rising operational costs, Microsoft is shifting GitHub Copilot to token-based billing, tightening rate limits, and has temporarily suspended new signups for individual plans.

A new benchmark from Zapier, called AutomationBench, highlights the current limitations of AI agents. The benchmark, which measures how well models perform real-world tasks like CRM updates and inbox follow-ups, found that no model has yet surpassed a 10% success rate.

Zapier (Workflow Automation) Logo
Zapier
4.6/5
#Industry News#Google#Anthropic#AI Competition#DeepMind#Sergey Brin#Claude
Máté Ribényi
AI Workflow & Efficiency Expert

Meet Máté Ribényi, Senior AI Workflow Auditor at testified.ai. With 15 years in business development and a background in IT project management, Máté audits productivity AI tools and workflow automations for real-world ROI.