How do VCs evaluate the potential of AI startups today compared to a few years ago and what are their challenges?
Venture capitalists (VCs) evaluate the potential of AI startups today with a more disciplined and practical approach than a few years ago, as you would expect.
Evaluation Differences Now and Then
Venture capitalists (VCs) evaluate the potential of AI startups today with a more disciplined and practical approach than a few years ago, as you would expect. This shift echoes lessons learned from earlier investment cycles, growing market dynamics, and the need for sustainable growth.
It got me thinking about this evolution and the differences between 2025 and now. How have they evolved, and what challenges are VCs facing with this fast-changing investment vertical?
1. The Transition from Hype to Practicality: A Key Shift in VC Evaluation
Then: (Pre-2025): VCs were heavily influenced by the hype around cutting-edge technology, often funding startups based on innovative algorithms or technical expertise without requiring clear business applications or ROI. Startups with buzzwords like "machine learning" or "neural networks" could secure funding even if their product-market fit was unclear[1][8].
Now: VCs prioritise startups solving specific, high-value problems with measurable outcomes. They demand proof of ROI, such as demonstrated cost savings or revenue generation. For example, startups automating hospital billing or fraud detection are more likely to secure funding than those offering vague promises of AI-powered transformation[1][3][8].
2. The Rise of Data as a Core Asset: A Fundamental Shift in VC Evaluation
Then: Investors focused on the uniqueness of AI models and technical complexity, often overlooking the quality and ownership of underlying datasets[1][5].
Data has become a critical differentiator. VCs assess a startup's data relevance, accuracy, and legality. Proprietary, high-quality datasets are now essential for building defensible AI solutions. Startups counting on open-source or public datasets encounter greater scrutiny due to concerns about scalability and future challengers[2][5].
3. Ethical AI and Regulatory Compliance: A New Standard in VC Evaluation
Then: Ethical considerations and regulatory compliance were often afterthoughts in VC evaluations. Startups were rarely asked about bias mitigation, privacy protocols, or compliance with laws like GDPR[1][8].
Now: Ethical AI practices and regulatory readiness are deal-breakers. VCs actively audit startups for transparency, fairness, and compliance with legal standards. Questions like "How do you mitigate algorithmic bias?" and "Can regulators shut you down?" are now central to due diligence processes[1][10].
4. Business Model Viability
Then: Startups were funded based on potential market size and growth projections rather than clear monetisation strategies. Early traction, such as free user adoption or pilot programs, was often enough to secure investment[1][8].
Now: VCs demand robust business models with clear revenue streams. Startups must illustrate how they plan to monetise their solutions—whether through subscription fees, API usage charges, or enterprise contracts—and show evidence of market demand[1][3][10].
5. Team Composition
Then: Teams with strong technical credentials (e.g., PhDs or ex-Big Tech engineers) were prioritised over operational expertise. A technically impressive team was often considered sufficient for securing funding[1].
Now: VCs look for balanced teams that combine technical excellence with business acumen. Founders must verify not only their ability to innovate but also their capacity to execute efficiently and scale [6][8].
6. Scalability and Infrastructure
Then: Scalability was often assumed rather than rigorously evaluated. Startups could secure funding without proving their ability to handle large-scale operations or diverse use cases[9].
Now: Scalability is a critical factor in VC evaluations. Investors assess whether a startup's technology can handle expanding datasets, computational demands, and broader market applications without compromising quality or margins[6][9].
7. Focus on Real-World Applications
Then: Generalist AI solutions aiming to disrupt broad industries (e.g., "AI for healthcare") were standard pitches that attracted funding despite lacking specificity[1].
Now: Task-specific solutions addressing niche but critical problems are preferred. Startups targeting enterprise customers with tailored solutions—such as compliance tools for fintech or supply chain optimisation—stand out due to their clear value propositions and focused execution[3][6].
What challenges do specialised VC funds face when investing in AI startups?
Specialised venture capital (VC) funds face various challenges when investing in AI startups, originating from the unique dynamics of the AI industry and broader market trends. These challenges include overvaluation, market saturation, scalability issues, ethical concerns, and regulatory complexities.
1. Overvaluation and ROI Concerns
The excitement surrounding AI has led to inflated valuations for many startups, often detached from proven revenue or profit-generating capabilities. Early-stage companies sometimes secure funding based on potential rather than tangible results, creating unrealistic expectations for long-term returns. This can result in reduced exit values and extended holding periods for investors who struggle to match lofty valuations with actual ROI[1][2].
2. Market Saturation
The AI sector is increasingly crowded, with numerous startups competing in popular niches like generative AI and machine learning automation. This oversaturation dilutes market share, lengthens profitability timeframes, and lessens the strategic value of particular investments. Specialised VC funds may need to navigate this competitive landscape by focusing on startups with distinct competitive advantages, such as proprietary algorithms or access to niche markets[1][3].
3. Capital-Intensive Business Models
AI startups frequently need substantial upfront investment in computational infrastructure, data acquisition, and specialised talent, which is getting harder to access and is at a premium. For example, training large language models can consume up to 80% of early-stage funding. Additionally, high salaries for AI engineers—averaging $300,000 annually—add to operational costs. These capital-intensive needs can sap the typical VC model and lead to challenges in securing follow-on funding at favourable valuations[2][3].
4. Scalability Challenges
AI technologies do not scale as easily as traditional software solutions due to dependencies on infrastructure, data consistency, and workforce expertise. Startups may excel in research and development but flounder with commercial scalability when adapting technology to varied real-world applications. Specialised VCs must evaluate a startup's ability to scale effectively while managing integration complexities[1][3].
5. Ethical and Regulatory Risks
Legal and ethical issues are becoming critical factors in AI investments. Concerns about algorithmic bias, privacy violations, deepfake proliferation, and copyright compliance are growing as governments introduce stricter regulations worldwide. Navigating these evolving frameworks requires startups to develop robust governance structures and ethical oversight mechanisms—areas that can be costly and time-consuming[6][7].
6. Winner-Take-All Dynamics
The AI industry is increasingly dominated by a few large players like OpenAI and Anthropic, which capture most of the profits due to their scale and resources. This "winner-take-all" dynamic poses risks for smaller startups that struggle to compete against well-funded incumbents or rely on commoditised technologies like cloud computing and semiconductors[2][3].
7. Economic Headwinds
Broader economic trends also impact AI investments. Weak earnings have made businesses wary about deploying large-scale AI solutions, while rapid technological advancements create hesitation about the longevity of current innovations. These factors contribute to a challenging fundraising environment for AI startups[7].
Strategies to Mitigate Challenges
To address these challenges:
Specialised VC funds are starting to diversify portfolios across underexplored sectors like renewable energy or edge computing[1][4].
They emphasise rigorous due diligence on scalability potential and ethical compliance[3][6].
Funds are shifting focus from pure innovation to practical applications with measurable outcomes[3][4].
What are the key factors driving the increase in VC investments in AI?
The rise in venture capital (VC) investments in artificial intelligence (AI) is driven by several key factors, reflecting AI technologies' transformative potential and widespread adoption across industries. Here are the primary drivers:
1. Economic Potential and Scalability
AI technologies offer exponential returns, making them highly attractive to investors. Startups working on foundational AI models or mission-critical applications—such as generative AI, industrial automation, and healthcare diagnostics—demonstrate scalability and global economic influence. For example, generative AI funding nearly doubled from $24 billion in 2023 to $45 billion in 2024, with projections indicating growth to $1.3 trillion over the next decade[18][19][20].
2. Industry Adoption Across Sectors
AI has become an embedded infrastructure in healthcare, finance, and logistics. For instance:
Healthcare startups are leveraging AI for drug discovery and personalised medicine.
Financial institutions are using AI for fraud detection and risk management.
Supply chains are optimizing operations with predictive analytics[19][20][24].
This widespread adoption underscores AI's ability to solve complex problems, driving investor confidence.
3. Strategic Focus on Monetizable Applications
The investment landscape is moving from foundational infrastructure to monetising AI applications. VCs are increasingly backing companies that integrate AI into their business models to deliver tangible outcomes, such as vertical-specific workflows or enterprise solutions that reduce friction and enhance efficiency[21][24].
4. Mega-Deals and High-Valuation Startups
Record-breaking funding rounds for prominent AI firms like OpenAI and Shield AI highlight the trend of VCs supporting large-scale projects capable of redefining industries. In 2025, startups securing $100M+ deals are concentrated in sectors like defence, healthcare, and SaaS, which promise profitability and societal impact[18][23].
5. Regulatory Developments
Governments worldwide are introducing frameworks or guidelines to regulate AI technologies, addressing data privacy, algorithmic bias, and security risks. While this creates compliance challenges for startups, it also signals maturity in the sector, attracting disciplined investments from VCs who value sustainable growth over hype[19][20].
6. Resurgence of IPOs
The growing IPO appetite for AI companies is another driver of VC interest. Major players like Databricks and CoreWeave appear to be preparing public offerings in 2025, echoing strong growth prospects and favourable market conditions that fuel optimism around AI investments[20].
7. Concentration on Established Leaders
Capital is increasingly flowing to dominant firms in the AI space, such as OpenAI and Anduril. This focus on established leaders suggests that investors prioritise proven business models over speculative ventures while selectively backing smaller startups with unique value propositions[24].
Final Thoughts
Today's VC evaluation criteria reflect a more mature AI investment landscape where practicality, defensibility, and sustainability precede hype and innovation. Startups must demonstrate technical prowess, ethical governance (adherence to ethical principles in AI development and deployment), scalable infrastructure, viable business models, and real-world impact to secure funding in 2025's competitive environment.
Specialised VC funds can navigate these challenges while capitalising on AI technologies' transformative potential by adopting disciplined investment strategies and aligning with sustainable business models.
These factors collectively explain why venture capitalists are doubling down on AI investments in 2025, recognising its transformative impact across industries and its potential to redefine economies globally.
References and Further Reading
[1] How DeepSeek Rewrote the Rules Venture Capitalists - Labellerr https://www.labellerr.com/blog/how-deepseek-rewrote-the-rules-venture-capitalists/
[2] VC's Framework for Evaluating an AI Startup's Tech Stack. | Sahil S https://www.linkedin.com/posts/sahilsr_vcs-framework-for-evaluating-an-ai-startups-activity-7274323044788035584-PhEF?_bhlid=16dc74d93eafa44bcfb9d58754a0e67a608d3257
[3] How to Be Investable AI Startup in 2025 - by Stepan Ikaev https://thecreatorsai.com/p/how-to-be-investable-ai-startup-in
[4] The 7 Secret Evaluation Criteria Venture Capitalists Use To Make ... https://thevcfactory.com/7-secret-evaluation-criteria/
[5] How VCs Evaluate AI Startups: The Frameworks You Need to Know https://www.linkedin.com/pulse/how-vcs-evaluate-ai-startups-frameworks-you-need-know-asthana-kzu5e
[6] AI-First vs. AI-Enabled Investment Framework - LinkedIn https://www.linkedin.com/pulse/ai-first-vs-ai-enabled-investment-framework-serhat-pala-7tewc
[7] What VC Investments Look Like in 2025 - Information Week https://www.informationweek.com/it-leadership/what-vc-investments-look-like-in-2025
[8] The State of the Funding Market for AI Companies: A 2024 - Mintz https://www.mintz.com/insights-center/viewpoints/2166/2025-03-10-state-funding-market-ai-companies-2024-2025-outlook
[9] Decoding The Evaluation Process Of Generative AI Companies By ... https://www.forbes.com/sites/josipamajic/2024/01/03/decoding-the-evaluation-process-of-generative-ai-companies-by-venture-capitalists/
[10] How to Invest in an AI Startup: What You Need to Know - Sembly AI https://www.sembly.ai/blog/investing-in-an-ai-startup/
[11] AI and Venture Capital: Redefining Investment Strategies in 2025 https://www.mahdlo.net/blog/ai-venture-capital-innovation
[12] [PDF] Navigating AI Investment Risks and Opportunities in Venture Capital https://meketa.com/wp-content/uploads/2024/11/MEKETA_Navigating-AI-Investment-Risks-and-Opportunities-in-Venture-Capital.pdf
[13] https://www.labellerr.com/blog/how-deepseek-rewrote-the-rules-venture-capitalists/
[14] A Reality Check for AI-Focused VCs: Deepseek's Disruption - LinkedIn https://www.linkedin.com/pulse/reality-check-ai-focused-vcs-deepseeks-disruption-taghashio-8gvbf
[15] AI craze getting funded by tech giants, distorting traditional VCs https://www.cnbc.com/2024/09/06/ai-craze-getting-funded-by-tech-giants-distorting-traditional-vcs.html
[16] AI investments in 2025: strategies VCs are using to navigate industry ... https://www.vestbee.com/blog/articles/ai-investments-in-2025-strategies-v-cs-are-using-to-navigate-industry-shifts
[17] Decoding AI Investment: Trends, Challenges, and Opportunities https://www.k4northwest.com/articles/decoding-ai-investment-trends-challenges-and-opportunities
[18] AI and Venture Capital: Redefining Investment Strategies in 2025 https://www.mahdlo.net/blog/ai-venture-capital-innovation
[19] The State of the Funding Market for AI Companies: A 2024 - Mintz https://www.mintz.com/insights-center/viewpoints/2166/2025-03-10-state-funding-market-ai-companies-2024-2025-outlook
[20] AI Investment Trends 2025: VC Funding, IPOs, and Regulatory Chall…https://natlawreview.com/article/state-funding-market-ai-companies-2024-2025-outlook
[21] Here are the types of AI companies enterprise VCs want to back in ... https://techcrunch.com/2025/01/20/here-are-the-types-of-ai-companies-enterprise-vcs-want-to-back-in-2025/
[22] What VC Investments Look Like in 2025 - Information Week https://www.informationweek.com/it-leadership/what-vc-investments-look-like-in-2025
[23] Venture Capital In 2025: Navigating The AI Hype And A Two-Speed ... https://gunungcapital.com/venture-capital-in-2025-navigating-the-ai-hype-and-a-two-speed-world/
[24] The AI Supercycle™ Report: Key Trends for AI Investors - VistaShares https://www.vistashares.com/the-ai-supercycle-report-key-trends-for-ai-investors-2/