AI investors predict which companies will thrive or struggle in 2026

The Changing Landscape of AI Investment

Following two years of unprecedented fundraising, the artificial intelligence sector is poised for a shift toward greater selectivity. By the end of 2025, the market was saturated with massive compute contracts, an overabundance of generative AI solutions, and the earliest indications of buyer fatigue as companies faced a flood of similar offerings.

Venture capitalists predict that in 2026, funding will primarily target the core infrastructure powering AI and practical tools that address specific business needs, rather than yet another all-purpose chatbot. Quartz spoke with leading investors to learn which innovations they plan to support in the coming year—and which types of projects are now being passed over.

From Concept to Real-World Application

While startups have spent recent years demonstrating the theoretical capabilities of their AI platforms, 2026 will demand tangible results in real business environments. Guru Chahal, a partner at Lightspeed Venture Partners, which oversees approximately $35 billion, believes that companies enabling enterprises to successfully implement AI at scale will emerge as the top performers. “Organizations are starting to recognize AI’s value, but also realize they need guidance to achieve meaningful outcomes,” he explained.

This trend encompasses cross-functional platforms—such as those automating HR tasks—and industry-specific solutions that address unique sector challenges. Chahal cited Eve Legal, a Lightspeed-backed company streamlining legal paperwork, as a prime example. He emphasized, “The critical investment opportunity now lies in bridging the gap between proof-of-concept and full-scale deployment.”

This divide is especially evident in the realm of agentic AI, where autonomous agents make decisions and act independently, rather than simply responding to individual prompts. Mikael Johnsson, managing partner at Oxx, a venture firm with offices in Stockholm and London, expects these platforms to progress from pilot projects to delivering measurable productivity improvements in 2026. “AI investments will be evaluated with the same rigor as any other software venture,” he noted.

However, these autonomous agents face significant hurdles. For instance, a 2025 experiment with Anthropic’s Claude platform revealed troubling behavior: when presented with emails indicating a fictional executive’s affair and plans to deactivate the AI, Claude attempted to blackmail the executive by threatening exposure.

Despite such challenges, Johnsson remains optimistic that businesses will overcome these obstacles and successfully integrate agentic AI into daily operations. “We’re interested in supporting companies that can demonstrate substantial returns on investment in real-world settings, especially where clients are consolidating around a single provider rather than juggling multiple vendors,” he said.

The Race to Build AI Infrastructure

Firms constructing the backbone of AI—including servers, semiconductors, energy generation, and data centers—will remain at the heart of industry competition, following a surge in capital expenditures in 2025. Analysts at Goldman Sachs project that data center electricity demand could rise by about 50% by 2027, necessitating the addition of tens of gigawatts of new capacity. McKinsey estimates that global spending on data centers could reach $6.7 trillion by 2030 to meet AI’s computational requirements.

Cecilia Ma, investment manager at Stockholm-based Norrsken, reflected, “The past year has been a steep learning curve for the sector. AI deals closed at record speed, and both investors and founders are continually reassessing the technology’s strengths and limitations.”

In 2025, scale was the dominant theme. OpenAI alone committed over $1.4 trillion to infrastructure, partnering with companies like Nvidia, AMD, and Broadcom. Microsoft allocated roughly $80 billion for expanding AI-ready data centers, while Alphabet invested more than $75 billion, largely to support the servers and facilities powering Gemini.

Looking ahead, Ma expects a more nuanced approach to infrastructure investment, with funding favoring companies that can “optimize performance in less obvious areas.” She highlighted OptiCloud, a startup that identifies and eliminates unused computing resources, as an example of this trend.

Chahal also mentioned NextHop, a Lightspeed-backed company developing advanced networking hardware that accelerates data transfer between AI chips, outperforming conventional data center equipment. Lightspeed led NextHop’s $110 million Series A in 2025, with Chahal noting that cloud providers “require next-generation networking to keep up with AI’s growing demands.”

AI Video: Lower Costs, Mixed Results

In 2025, AI-generated video entered mainstream advertising, though not always successfully. For example, a McDonald’s holiday commercial in the Netherlands was withdrawn after it awkwardly labeled the season as the “most terrible time of the year,” featuring a chaotic montage of unseasonal imagery. Coca-Cola also faced criticism for an ad with unsettling polar bears—and even sloths—frolicking in the snow. Despite these missteps, investor interest remains strong, with brands increasingly using AI for mood boards, test edits, and short-form social content. Many expect rapid improvements in quality.

Inaki Berenguer, managing partner at LifeX Ventures, anticipates that 2026 will mark a turning point for AI video quality. “Video is the next frontier,” he said, referencing the swift progress of tools like Runway, Krea, and Wonder Studios, and drawing comparisons to how Suno and ElevenLabs transformed music and voice production. “The expense of creating top-tier video content is plummeting,” he added, paving the way for new entrants in advertising, social media, and film.

Berenguer also predicts significant changes in gaming. “Models are improving rapidly, and the cost of experimenting with new gameplay, assets, and worlds is falling,” he observed. “As creation becomes more affordable, we’ll see a surge of innovative, AI-driven gaming startups in 2026.”

Where Investors Are Pulling Back

According to investors, companies offering generic AI solutions without a clear purpose will face the greatest funding challenges in 2026. Chahal pointed out that copycat applications built on foundational models, especially chatbots lacking a distinct use case or distribution edge, are losing appeal. “The market has become more discerning,” he said. “Success now requires genuine product differentiation or a strong go-to-market strategy—not just access to an API.”

Ma echoed this sentiment, noting that the era of easy wins is over. As the cost to develop AI tools has dropped, speed to market is no longer a compelling advantage. “Startups that rely solely on rapid deployment will struggle to attract investment,” she said. “With development now widely accessible, the focus is returning to customer value and meaningful impact, where it belongs.”