Artificial intelligence continues to dominate headlines across the U.S. economy as investors, regulators, and businesses grapple with its potential. Recent reports show billions pouring into the sector even as questions arise about governance and safety. California’s veto of a landmark AI bill has sparked fresh debate, while SoftBank’s $500 million investment into OpenAI highlights investor confidence. As interest in generative AI grows, the next phase of development and adoption is beginning to take shape—unevenly, but with clear momentu
Generative AI Investment Surges Despite Ongoing Market Challenges
Spending on generative AI (GenAI) is climbing at a pace that outstrips nearly every other tech segment. According to new PitchBook data, U.S. firms spent $7.5 billion on GenAI software in 2023. By August 2024, that number had more than doubled to $17 billion. Analysts projected it would reach $32.4 billion by 2025.
Industry insiders reported that venture capital activity followed a similar trajectory. In 2022, there were 581 GenAI deals worth $9.3 billion. By 2023, the number had jumped to 877 deals totaling $26 billion. As of late August 2024, 508 deals had already raised $23.9 billion. These figures signal that GenAI investment is not only accelerating but doing so while other tech sectors face tightening capital conditions.
SoftBank’s recent commitment of $500 million to OpenAI’s funding round, valuing the company at $150 billion pre-money, further reflects strong institutional faith in the foundational AI space. Thrive Capital also pledged $1 billion, with Microsoft and other firms participating. These developments suggest that well-capitalized players continue to drive funding at the model layer, even as application-level startups struggle to find traction.
Read also on Claude AI vs. ChatGPT in 2025: Which AI Assistant Is Right for You?
Enterprise Adoption Lags Amid High Costs and Complex Sales Cycles
Despite investment enthusiasm, adoption of GenAI tools in enterprise settings has proven slower than expected. Business lawyers working with AI firms report that adoption barriers persist across industries. High computing costs, limited data availability, and security concerns remain core challenges, especially when trying to integrate AI into legacy systems.
One key factor is the absence of a “plug-and-play” solution. AI vendors must often engage in detailed discovery sessions with clients to identify business-specific use cases. Analysts noted that customers rarely approach AI vendors with ready-made deployment plans. Instead, the process involves demonstrations, proofs of concept, and pilots—contributing to a lengthy sales cycle.
Post-zero interest rate policy (ZIRP) conditions have further complicated this environment. Startups that raised capital during the ZIRP era now face difficulty scaling in a market demanding fast returns. According to PitchBook analysts, investors are now scrutinizing application-layer startups for proof of commercial viability, a shift that has added pressure to young firms with long development cycles and slow client conversions.
Varied Impact of AI Across Industries: Fintech, Crypto, and More
suggested that GenAI could significantly transform crypto by supporting smart contract development and auditing. By 2024, the impact had been more muted than anticipated. Users remained reliant on familiar tools, and the need for human oversight in coding created friction.
Experts pointed out that adoption inertia was not just about technology—it was also about behavior. Many developers were simply more comfortable using legacy platforms and reluctant to embrace new AI workflows without guarantees of security and control.
In fintech, a slower pace of GenAI adoption had been forecast due to regulatory pressures and long R&D cycles. That forecast proved partially accurate. While large financial institutions continued to act with caution, fintech startups moved more quickly. By 2024, smaller firms had begun rolling out GenAI-based tools for customer service, internal documentation, and risk analysis. Still, full-scale deployment remained limited by compliance needs and internal audit standards.
Regulatory Uncertainty Adds Friction to AI Momentum
Amid mounting excitement over AI’s economic promise, the U.S. regulatory landscape remains murky. In California, Senate Bill 1047 sought to establish rules requiring developers of large language models to exercise “reasonable care” and provide human overrides in the event of dangerous behavior. The proposed legislation aimed to make AI safer and more accountable.
The Gavin Newsom, the governor of California, ended up vetoing the bill. He proved that it was specific to the biggest, most costly models and did not look at circumstances of applications. This move received criticism among safety advocates and was hardly welcomed by startups and many investors. The leaders of the venture capital were worried that the bill would impose unreasonable burdens on the new firms, a move that would protect incumbents.
Read also on Top AI Trends in Data Analysis to Watch in 2025
Few regulatory proposals are being made at the federal level. Although legislators are also worried about AI’s potential dangers, non-exhaustive laws are not advancing. Those who were observing it said that in the absence of guardrails, the path of the technology was still likely to be determined by market forces rather than legal limits.
The Role of Foundation Model Startups Versus Application Layer Firms
The largest share of investor interest has gone to the companies that are constructing foundation models, including OpenAI, Anthropic, Hugging Face, and Mistral, in the present setting. Their efforts in mainstream LLMs have an equal fascination of headlines and venture capital.
Application-layer firms, in contrast, face more obstacles. These businesses must tailor solutions to individual use cases, which often means engaging with customers on a case-by-case basis. The “use-case first” approach makes rapid scaling difficult. Unlike foundational model firms, application-level startups can’t rely on a single product for mass distribution.
Consequently, startups in the application business have to find an equilibrium between innovativeness and the necessity of immediate outcome. Investors are also less and less willing to finance length exploratory periods without a demonstration of market demand. Attorneys in the AI world have been reporting that this operating environment has prompted many companies to make course changes, scale back, or inquire about being joined by others in order to stay afloat.
Broader Outlook: GenAI Still in Early Stages of Adoption
After all the investments, GenAI is not yet on its adoption curve. The analysts hold that the technology does not have full potential tapped in it. Numerous companies are struggling to determine where GenAI tools can be used in their operations, and in which departments or processes they can effect positive value.
Experts agree that adoption will take time. The technology is powerful, but not always easy to implement. Companies must navigate integration challenges, employee retraining, and internal resistance. Legal questions around data privacy, intellectual property, and liability also remain unresolved.
Nevertheless, there is force of momentum. By mid-2024, venture capital and product drops, as well as interest among enterprises, indicate GenAI is not going anywhere. The way to go is not smooth, yet the following year is sure to continue development and experimentation.
Read also on Building Trust: Responsible AI Development Guidelines for US Innovators
Broader Outlook: GenAI Still in Early Stages of Adoption
After all the investments, GenAI is not yet on its adoption curve. The analysts hold that the technology does not have full potential tapped in it. Numerous companies are struggling to determine where GenAI tools can be used in their operations, and in which departments or processes they can effect positive value.
Experts agree that adoption will take time. The technology is powerful, but not always easy to implement. Companies must navigate integration challenges, employee retraining, and internal resistance. Legal questions around data privacy, intellectual property, and liability also remain unresolved.
Nevertheless, there is force of momentum. By mid-2024, venture capital and product drops, as well as interest among enterprises, indicate GenAI is not going anywhere. The way to go is not smooth, yet the following year is sure to continue development and experimentation.
Read also on Building Trust: Responsible AI Development Guidelines for US Innovators
FAQs
Why are investors pouring money into generative AI despite economic uncertainty?
Generative AI is viewed as a transformative technology with broad applications, attracting major investments from firms like SoftBank and Microsoft despite tighter capital markets.
What challenges are slowing the adoption of generative AI in U.S. enterprises?
High computing costs, system complexity, and the need for customized use cases make implementation slow and resource-intensive for most businesses.
How did California’s veto of Senate Bill 1047 impact AI regulation in the U.S.?
Governor Newsom’s veto halted a key effort to regulate large AI models, raising concerns about the absence of clear safety and accountability standards nationwide.
Which industries are adopting generative AI the fastest in the U.S.?
Fintech startups are moving quickly with GenAI integration for customer support and documentation, while crypto and traditional banks remain cautious.
What’s the difference between foundational AI model firms and application-layer startups?
Foundational model firms like OpenAI build general-purpose AI tools, while application startups create tailored solutions that often face longer sales cycles and adoption hurdles.
