Binance’s Founder CZ Calls for AI Companion Integration In Crypto

Calvin James

Binance founder Changpeng Zhao (CZ) has identified artificial intelligence companions as an up-and-coming market, calling upon its implementation into cryptocurrency platforms. Zhao called AI companions a large and expanding discipline in a social media publication. He pointed out that users are ready to pay for such technologies, which is why they can be a potential fit in crypto-based ecosystems. Zhao said, Let’s bring them to crypto, or vice versa also emphasized his proposal to combine digital assets and artificial intelligence applications.

Source: X

His message indicates the tendency to broaden the scope of crypto outside of trading and payments to the places where the user experience and engagement with new technologies meet each other. AI companions are increasingly monetized, frequently designed to offer conversational, educational, or entertainment utility, and Zhao proposed that blockchain infrastructure might have a role in enabling ownership, transactions, and digital identities attached to them.

Coinbase Reports Surge in AI-Generated Code

Coinbase CEO Brian Armstrong introduced a concrete operational viewpoint to the discussion. In his quote, he unveiled that around 40 % of the daily written code at Coinbase is now generated with the help of AI. He has stated that he wants to increase this figure to more than 50 % by October 2025.

Armstrong explained that AI is very useful in software development, but not every business field would be appropriate to use it. He stressed the need to have human control, implying that it must obviously be revised and interpreted. In his view, Coinbase is taking a cautious stance, responsibly implementing AI-generated code and harmonizing it with industry standards regarding security and quality.

Source: X

In addition to the statement made by Armstrong, internal data at Coinbase indicate a steep rise of AI-generated lines of code (LOC). A table accompanying his remarks follows the share of AI-generated code between April and August 2025.

By April 1, 2025, Coinbase documented slightly more than 15 % of its daily code output done by AI systems. This percentage had reached 20 % by May 2025 and was steadily adopted by teams. Another threshold was June, when the levels were approaching 30%. The information reveals that in July, the data is fluctuating but sums it up by about 31 to 32%, after which it rapidly increases in August. By early August 2025, more than 40% of the code was generated via AI, as Armstrong expects in October.

As can be seen in the chart the growth is steady in the four months period and the gain is almost 25% points within less than four months. It means that Coinbase does not just implement AI tools at scale but also extends the implementation of AI in various engineering activities.

Coinbase’s AI Integration Timeline

The timeline gives an understanding of how quickly AI will be adopted in Coinbase. The amount of AI-generated code percentage rose steadily and steadily between April and the beginning of June. It was then succeeded by a plateau in July, probably due to the stabilization and optimization of workflows. August can be seen as a sharp increase in the growth and the levels went up as high as 32% towards the end of July and above 42 % towards the end of the first week of August.

This is an increase of 10% in one month. The 50 % mark of the growth rate identified by Armstrong might be met earlier than the projected figure in October in case of sustained growth rate. The statistics of the company prove the fact that the development based on AI changed experimental use to the essential part of its engineering activities.

The remarks made by Armstrong emphasized that AI does not negate the existence of human engineers but rather compliments them. He admitted that AI is not applicable to all operations. Critical points, including security systems, compliance modules, and financial transactions guidelines, are to be strongly human-reviewed. But when it comes to repetitive coding work, bug detection and optimization, AI has already been found useful.

The review and understanding requirement show that Coinbase is integrating human-in-the-loop defenses. They are meant to curb dangers linked to AI-based mistakes, hallucinations, or inefficiencies. Armstrong positioned the adoption as a responsible strategy, in line with bigger standards of corporate governance in the crypto and tech industries.

Industry-Wide Implications of AI in Crypto

The two parallel statements by Zhao and Armstrong focus on two different but related uses of AI in the crypto industry. On the one hand, Zhao focuses on consumer uses of the so-called AI companions that might be incorporated into blockchain infrastructure, which may be backed by tokens or decentralized infrastructure. On the other hand, Armstrong concentrates on backend efficiency- using AI to build software to increase product rollout and enhance scalability.

Coinbase data indicates that AI usage is quantifiable, as almost half of its code output is now based on automated code. This degree of integration depicts that artificial intelligence is transitioning out of pilot programs to mainstream use in large cryptocurrency companies.

The 40% mark is a big improvement over 2024, when AI was at its early stages of use. Although Coinbase has not released the historical data before April 2025 in the chart, commentary in the industry indicated that AI input on coding work was less than 10 % at the beginning of 2024. It implies that the AI-generated output of Coinbase has increased four times within less than 18 months.

This acceleration in May–August 2025 is in line with more general progress in coding-specific AI tools, such as large language models trained with software development in mind. The uptake curve in Coinbase reflects the industry trends in which AI-assisted programming has become part of the workflow of developers.

The trend line reflects the fact that Coinbase is ramping up to majority AI-generated code. Armstrong has a chance of reaching the October target of crossing 50%, given that the company had already crossed 40 % in August. A linear projection of the rate of growth indicates that prior to the stated time, Coinbase might hit more than half of its code being written by AI.

Zhao, in the meantime, is calling AI companions, which is a different area of growth, that makes consumers more than increasing business efficiency. This is one of the areas where blockchain may be used to facilitate ownership validation, micropayments, and decentralized hosting of AI-driven platforms.

Source