International Edition
Latest News
Technology

Alphabet shares dip following internal skepticism over Gemini 4 AI performance

Alphabet Inc. shares dipped toward the end of a recent trading session, trimming earlier gains to roughly 0.5% following a Bloomberg report detailing internal employee skepticism over the performance of the company's newest artificial intelligence model, Gemini 4.…

Alphabet shares dip following internal skepticism over Gemini 4 AI performance

Alphabet Inc. shares dipped toward the end of a recent trading session, trimming earlier gains to roughly 0.5% following a Bloomberg report detailing internal employee skepticism over the performance of the company’s newest artificial intelligence model, Gemini 4. The report exposed a sharp divide between strong industrial benchmark scores and practical utility in everyday developer tasks.

Benchmark Success Versus Practical Coding Realities

While Gemini 4 performs well on standardized industry benchmarks used to measure overall AI effectiveness, internal users reported struggles when applying the model to real-world software engineering tasks. Several employees familiar with the project noted that the model faced difficulties handling complex coding assignments.

Diverging Internal Views on Competitive Standing

Internal opinions regarding the technology remained split. Some staff members argued that competing offerings, such as Anthropic’s Fable and OpenAI’s Astra, were progressing at a faster pace. Conversely, other employees maintained that Gemini 4 successfully matched the capabilities of leading rival AI labs, creating what one source described as a major internal consensus that Google remains at the forefront of the sector.

Alphabet shares dip following internal skepticism over Gemini 4 AI performance

Leadership Pushback and Altered Release Timelines

Google strongly disputed the characterization that Gemini versions underperform in practical applications like coding. Company representatives pointed to statements made by Koray Kavukcuoglu, head of Google DeepMind, who expressed strong optimism regarding the model’s performance trajectory.

Internal planning also shifted prior to the release cycle. Reports indicated that Google previously considered launching an intermediate version named Gemini 3.5 Pro in June before ultimately canceling those plans.

Foundational Integration Across Billion-User Products

Despite the internal debate, variants of the Gemini architecture serve as furnaces for nearly every consumer product sold by Google. The technology powers AI-driven features across Search, Maps, Gmail, and Chrome, each of which serves more than one billion active users globally.

About the author: Anika Shah - Technology

MSc in Computer Science, senior reporter. Anika focuses on AI ethics, cybersecurity, and emerging hardware—frequently moderating panels at CES and Web Summit. “Anika Shah decodes tech breakthroughs and startup disruption shaping tomorrow’s digital landscape.”