Google CEO Sundar Pichai on the one AI race Google is ‘losing’ to Anthropic and OpenAI
Google CEO Sundar Pichai admitted that Google is lagging behind competitors Anthropic and OpenAI in the field of 'agentic coding,' a critical area of artificial intelligence that involves complex coding tasks, tool integration, and long-term project execution. During an interview on the Hard Fork podcast, Pichai stated that while Google's models excel in areas such as text processing, multimodality, and reasoning, they fall short in agentic coding capabilities. This gap is attributed to Google's lack of a dedicated environment for developers, which hindered data flow and innovation in this domain. Pichai expressed confidence in Google's ability to catch up, highlighting new developments like Antigravity 2.0 and Gemini 3.5 Flash, which aim to address these shortcomings. The competition in agentic coding is seen as crucial for capturing developer attention and driving enterprise adoption, with significant financial stakes involved.
Microsoft’s aggressive cost-cutting strategy has escalated to include its key AI partners, OpenAI and Anthropic, as the tech giant seeks to reduce reliance on external models and bolster its own internally developed AI systems. According to reports, Microsoft is gradually replacing AI functionalities previously powered by OpenAI’s GPT and Anthropic’s Claude with its newly developed “MAI” models. These changes are evident in widely used software such as Microsoft Excel and Outlook, where AI-driven tasks, like generating emails or analyzing data, are increasingly handled by Microsoft’s in-house models. This shift marks a strategic pivot toward greater self-sufficiency in AI development, allowing Microsoft to control costs and improve efficiency. The decision follows a public warning issued by Mustafa Suleyman, head of Microsoft’s AI division, during a June meeting. Suleyman emphasized the need to minimize financial dependency on external AI providers, particularly Anthropic, whose models are integral to many of Microsoft’s services. He highlighted the high computational costs associated with AI operations, noting that each interaction involving AI assistants consumes vast amounts of “tokens”, the fundamental units of AI processing. With Microsoft’s Copilot consuming these tokens at an alarming rate, the company faces pressure to find more sustainable alternatives. Suleyman’s remarks underscored the urgency of developing proprietary solutions to avoid future price volatility tied to third-party vendors. In response to these challenges, Microsoft unveiled several new in-house AI models at its June Build developer conference. One of these models, designed for coding tasks, is claimed to perform comparably to Anthropic’s earlier Opus 4.6 model but at a significantly reduced operational cost. Suleyman expressed confidence in the potential of these models, stating that reducing expenses with external providers was a central objective. The integration of these models extends beyond basic productivity tools, with plans to incorporate them into GitHub Copilot for developers and Microsoft Teams for video conferencing. As these models become more embedded in Microsoft’s ecosystem, they are expected to play a growing role in shaping the company’s AI landscape. Meanwhile, Google CEO Sundar Pichai admitted that his company is lagging behind competitors in the critical area of agentic coding, a domain dominated by Anthropic’s Claude Code and OpenAI’s Codex. During a conversation on the Hard Fork podcast, Pichai acknowledged that Google’s models excel in areas like text generation and multimodal processing but fall short in handling complex, long-term coding tasks. He attributed this gap to a lack of direct access to developer workflows, which are essential for training models tailored to specific use cases. Pichai emphasized that Google’s approach is not a failure of focus but rather a challenge of scale and context, as the company operates across numerous domains simultaneously. The competition in agentic coding has intensified, driven by the lucrative potential of AI-powered development tools. Anthropic’s success with Claude Code and OpenAI’s pivot to enterprise-focused Codex have positioned them as leaders in this space. Analysts predict the AI code tools market will expand dramatically, reaching nearly $30 billion by 2031. Despite these challenges, Pichai remains optimistic, pointing to recent advancements such as Google’s Antigravity 2.0 and Gemini 3.5 Flash. These updates aim to address the coding gap by offering faster performance and enhanced multitasking capabilities. However, early adopters have raised concerns about usage limitations and scalability issues, prompting Google to adjust quotas and refine its offerings. Internally, Google is taking steps to close the coding gap, with initiatives led by figures like Sergey Brin and DeepMind’s Koray Kavukcuoglu. Internal metrics suggest progress, including a doubling of token usage within Google’s systems each week. Yet, the company still trails behind Anthropic in AI-assisted coding, with the latter achieving near-complete automation. Google’s efforts to introduce new pricing tiers and optimize its models reflect a broader strategy to enhance competitiveness. As the AI landscape continues to evolve, both Microsoft and Google face mounting pressure to innovate rapidly and secure a dominant position in the ever-expanding world of artificial intelligence.
How each side covered it
The same event, grouped by the political lean of the outlets covering it.
progressive
center
conservative
★
How each side covered it
Support independent, bias-aware news and unlock the social pulse, community voting, and your personalized For You feed.
The United States has lifted restrictions on several AI models, including Fable 5 and Mythos 5, allowing them to resume operations. Anthropic, one of the companies affected by these restrictions, has restored access to its AI models. The decision comes after a period of regulatory scrutiny and potential compliance issues. This development marks a shift in U.S. policy toward AI regulation, potentially impacting both domestic and international AI development.
Bias read (Center): The article reports on a policy change related to AI regulations, which is a politically charged issue. However, the tone remains neutral, presenting the facts without overtly favoring any particular political stance. The focus is on the action taken by the U.S. government rather than advocating for
Why these scores (Factual 85 · Objective 90): Accurately reports on US lifting restrictions and Anthropic's restoration of access. Consistent with cross-source consensus. Balanced and neutral tone.
Times of IndiaIndependentCenterFactual 85Objective 707 days ago
Microsoft is reducing its reliance on external AI partners like OpenAI and Anthropic by integrating its own internally developed AI models into key products such as Excel, Outlook, and GitHub Copilot. This shift comes after significant layoffs and a broader cost-cutting strategy. According to Bloomberg, Microsoft's 'MAI' models are handling tens of thousands of AI prompts weekly, previously managed by third-party technologies. Microsoft's AI division head, Mustafa Suleyman, emphasized the need to decrease financial dependency on external AI providers, particularly as current discounted rates for AI computing power from OpenAI may not be sustainable. Microsoft recently introduced seven new in-house AI models, including one capable of matching Anthropic's advanced coding model at a lower cost. These internal models are expected to expand into more core products like Microsoft Teams.
Bias read (Center): The article presents factual information about Microsoft's strategic decisions regarding AI development and cost reduction. It does not exhibit overtly biased language, one-sided sourcing, or omission of context. The content focuses on corporate strategy rather than political ideology or controversy
Why these scores (Factual 85 · Objective 70): Factuality is high as the article aligns with the cross-source consensus about Microsoft reducing reliance on external AI models. Objectivity is lower due to speculative language such as 'dumping models' and emphasis on internal cost-cutting benefits without balancing potential negative impacts on p
Times of IndiaIndependentCenterFactual 80Objective 853 days ago
Google CEO Sundar Pichai admitted that Google is lagging behind competitors Anthropic and OpenAI in the field of 'agentic coding,' a critical area of artificial intelligence that involves complex coding tasks, tool integration, and long-term project execution. During an interview on the Hard Fork podcast, Pichai stated that while Google's models excel in areas such as text processing, multimodality, and reasoning, they fall short in agentic coding capabilities. This gap is attributed to Google's lack of a dedicated environment for developers, which hindered data flow and innovation in this domain. Pichai expressed confidence in Google's ability to catch up, highlighting new developments like Antigravity 2.0 and Gemini 3.5 Flash, which aim to address these shortcomings. The competition in agentic coding is seen as crucial for capturing developer attention and driving enterprise adoption, with significant financial stakes involved.
Bias read (Center): The article presents a balanced view of the competitive landscape in AI development, focusing on technical challenges and strategic responses rather than taking a clear ideological stance. It reports on corporate strategies and technological advancements without overtly favoring any particular group
Why these scores (Factual 80 · Objective 85): Factuality is strong with direct quotes from Pichai and analysis of industry trends. Objectivity is high as it presents both challenges and optimism without bias, maintaining a neutral tone.
The Indian ExpressIndependentCenterFactual 75Objective 8010 days ago
The article reports that OpenAI might reduce its AI subscription prices due to increasing competition from Anthropic. This development comes amid growing market pressures as Anthropic challenges OpenAI's position in the AI industry. The report suggests that price adjustments could be part of OpenAI's strategy to maintain competitiveness. However, the article does not provide specific details about the extent of potential price cuts or the timeline for such changes.
Bias read (Center): The article presents information about corporate strategy in the AI sector without overtly favoring any particular political ideology. It focuses on business decisions rather than political stances, though the broader implications of AI regulation and market dynamics could intersect with policy. The
Why these scores (Factual 75 · Objective 80): Factuality is moderate as it reports on regulatory changes without additional context. Objectivity is good as it presents the restoration of access without taking sides or introducing unnecessary commentary.
Business StandardIndependent🔒CenterFactual 60Objective 5512 days ago
The article reports that OpenAI is considering offering the U.S. government a 5% equity stake in the company, amid discussions about potential federal involvement in artificial intelligence development. This comes as President Donald Trump has been contemplating increasing the government's ownership interest in AI-related technologies. The piece highlights the growing intersection between private technology firms and governmental oversight in the field of artificial intelligence.
Bias read (Center): The article presents information about both OpenAI's potential offer to the U.S. government and Trump's consideration of increased public involvement in AI without overtly favoring either side. It focuses on the discussion rather than taking a clear ideological stance on the matter.
Why these scores (Factual 60 · Objective 55): Factuality is low as the article introduces new information about a potential government stake without evidence. Objectivity is poor due to lack of balance and focus on political angles without contextual depth.
★
Keep the news honest.
ObjectiveNews is reader-funded and ad-free — we show you the bias instead of hiding it. Support independent journalism for €5/month.