AI Goes Mainstream: Goldman Rolls Out Assistant, Decagon Hits $1.5B & Google Faces Scrutiny

Artificial intelligence (AI) has moved from a tech trend to an everyday tool. It’s now helping banks run smoother, changing how companies talk to their customers, and even reshaping how online content is used behind the scenes.

However, as AI becomes more common, the questions around how it’s built and who it impacts.

Let’s take a look at three recent incidents that show just how much the AI world is shifting and what it means for the people and industries it touches. 
 

Goldman Sachs Rolls Out Firm-Wide AI Assistant

Goldman Sachs has rolled out a new AI assistant across the company, aiming to make day-to-day work easier and more efficient for its global workforce.

After testing it with around 10,000 employees, the bank is now making the tool available firmwide.

In an internal memo from Chief Information Officer Marco Argenti, the GS AI Assistant is designed to streamline a variety of tasks for Goldman employees. The memo specifies the tool will assist with "summarizing complex documents and drafting initial content to performing data analysis."


With this full-scale rollout, Goldman Sachs joins a growing number of major banks using AI to boost their operations and support everyday employee tasks. Citigroup, for instance, utilizes Citi Assist for internal policy searches and Citi Stylus for document summarization and comparison.

Morgan Stanley has deployed a chatbot to aid financial advisors in client interactions, while Bank of America's virtual assistant, Erica, focuses on assisting retail clients with day-to-day transactions.

The rollout is part of a broader trend across the finance world, where banks are increasingly turning to AI to stay competitive and keep up with the fast pace of change in the industry.

Beyond boosting productivity in boardrooms, AI tools are also attracting massive investments from VCs betting big on the future of customer service.
 

Decagon Achieves Unicorn Status with $131 Million Funding Round

San Francisco-based Decagon, a fast-growing player in the customer service tech space, announced on Monday that it has raised $131 million in its latest round of funding.

The latest funding brings Decagon’s valuation to an impressive $1.5 billion, officially giving it “unicorn” status in the crowded AI space. The round was led by well-known venture capital firms Accel and Andreessen Horowitz.

Existing investors A*, Bain Capital Ventures, and BOND also participated, alongside new contributors Avra, Forerunner, and Ribbit Capital.


This major investment comes less than a year after Decagon’s Series B round, where it raised $65 million at a $650 million valuation, highlighting just how quickly investor confidence in the company has grown.

Decagon now stands as one of the most highly valued AI startups dedicated to customer support solutions, competing with industry titans like Salesforce and agile startups such as Sierra, led by former Salesforce co-CEO Bret Taylor.

Sierra raised $175 million in October, reaching a $4.5 billion valuation. This growing investment in application-focused AI reflects a broader trend, with investors shifting from costly foundational models to practical tools that deliver steady revenue, especially since the rise of OpenAI’s ChatGPT.

Decagon plans to use the new funding to grow its team and expand into more enterprise markets, according to CEO Jesse Zhang. The company already works with a range of clients, including Hertz, Duolingo, Eventbrite, and Chime.

Zhang highlighted the tangible benefits for clients, citing financial services company Chime's experience, which saw a 60% reduction in contact center costs and a doubling of its net promoter score.

"When AI can take action and solve things, customers can get what they want much faster and more consistently," Zhang stated, emphasizing the efficiency and reliability offered by their AI solutions.

Decagon offers both text-based AI support, like chat and email, and advanced voice agents that can manage full customer service calls. It also provides tools that let businesses build their own custom AI support solutions.

As investment in AI grows, so do the ethical and legal questions, especially when it comes to how AI is trained. A recent spotlight on Google’s practices reveals the complex and often unseen dynamics behind AI development.
 

YouTube Content Fuels Google's AI Ambitions, Creators Largely Unaware

In a move that’s sparking fresh debate over digital rights and data use, it’s been revealed that Google is using content from its massive YouTube library to train its advanced AI models, Gemini and Veo 3.

While tech leaders have acknowledged the practice, many of the creators whose content is being used remain largely unaware that their work is helping fuel these AI developments.

With an estimated 20 billion videos, YouTube represents an unparalleled data source. Google, the platform's owner, has confirmed its use of a "subset of videos" for AI training, stating, "We've always used YouTube content to make our products better, and this hasn't changed with the advent of AI." The company also claims to honor specific agreements with creators and media entities.


However, many industry experts point to a major gap in awareness. Most creators and companies don’t realize their content is being used for AI training. What’s more, there’s currently no way for individuals to opt out of having their work used this way by Google.

Even just 1% of YouTube’s content amounts to a massive 2.3 billion minutes of training data, far surpassing the size of datasets used by many other AI models.

The implications are even more striking in light of Google’s recent launch of Veo 3, a video model that can generate highly realistic clips. Many observers note the irony—human-created content is now helping train AI tools that could eventually compete with those same creators or threaten their sources of income.

While creators can choose to opt out of third-party AI training by companies like Amazon and NVIDIA, there’s still no such option when it comes to Google’s own internal AI training—an omission that hasn’t gone unnoticed.

This situation highlights the shifting dynamics of digital content ownership and the growing demand for greater transparency as AI continues to reshape how creative work is used and valued.

Do you think people should know when their content is used to train AI? Do companies need to be more transparent about AI training?

Let us know your thoughts in the comments below!