Technology investment outlook 2027: AI moves from promise to proof

Pamela Hegarty, Co-Portfolio Manager of the Disruptive Technology Fund

Derek Glynn, Co-Portfolio Manager of the Disruptive Technology Fund

  • AI shifts from promise to measurable return on investment: 2027 should see AI move further from experimentation toward monetisation, productivity gains, and demonstrable returns, particularly across cloud and enterprise applications
  • AI infrastructure spending remains strong: Demand for GPUs, custom silicon, optical networking, data center power, and other infrastructure should remain robust, with supply constraints persisting in several areas
  • A constructive outlook balanced by rising risks: Technology valuations remain broadly supportive, but higher interest rates, debt-financed AI capex, geopolitics, and AI safety concerns could create periods of volatility

After several years of rapid innovation, the year ahead could mark the point at which artificial intelligence’s economic impact becomes increasingly visible.

The conversation is likely to shift from technological possibility to real-world adoption, monetisation, and return on investment.

Personal AI agents, enterprise automation, physical AI, and cloud infrastructure are all moving from experimentation toward broader deployment at scale.

The companies best positioned to monetise new products and distribute this innovation should be the primary beneficiaries of the AI cycle’s next phase.

Personal AI agents go mainstream

2027 could be the year that agentic AI transitions from an emerging technology to a mainstream consumer product.

Rather than simply answering questions, personal AI agents will increasingly perform tasks on a user’s behalf. Common use cases could include prioritising email messages, booking travel, scheduling appointments, managing subscriptions, and coordinating household services.

Improvements in reasoning, memory, and tool use should allow these agents to operate across applications and websites with minimal user input.

Adoption will likely be driven less by model breakthroughs and more by convenience, reliability, and integration into products consumers already use. Over time, personal AI agents could become a new interface layer for the internet.

AI moves into core enterprise workflows

Enterprise AI use cases will expand beyond coding assistants and customer service into agentic capabilities that augment core business workflows.

Organisations will increasingly deploy AI across functions such as finance, sales, human resources, and supply chain management. The next stage of the AI cycle will be defined not only by model capabilities, but also by the ability to deliver useful intelligence at an attractive cost.

Smaller models are likely to handle routine tasks, while larger models are reserved for more complex reasoning. Increasingly sophisticated orchestration software will manage these interactions behind the scenes, making AI more seamless for end users.

As adoption broadens, the focus should shift from experimentation to measurable productivity gains.

AI safety and cybersecurity remain top of mind

The AI safety debate will continue in 2027. Many industry experts fear a major cybersecurity breach or biohazard attack by bad actors using AI models. It is increasingly clear that cybersecurity, governance, and observability software will play a critical role going forward. 

Microsoft co-founder Bill Gates and others have advocated for deterministic supervisory software monitoring each AI model to build in policy guardrails for governance, compliance, and security.1

Other suggestions by industry experts include third-party auditing of training workloads and independent testing of models before public release. There is a growing consensus that governments need to regulate the industry, but also significant disagreement over how such regulations should be formulated.

Physical AI and edge computing build momentum

Next year could be a pivotal period for physical AI and edge computing – where data processing and storage is physically closer to the data source – as advances in AI begin translating into real-world automation.

Adoption is likely to accelerate first in structured environments such as factories, warehouses, and logistics networks, where intelligent robots can improve productivity and help address labour shortages.

Likewise, AI-enabled wearables, such as smart glasses, could emerge as an important new computing category by bringing contextual intelligence into everyday activities.

Humanoid robots may also make meaningful technical progress, but large-scale adoption likely remains several years away. Taken together, 2027 may represent the beginning of a new phase in which AI expands beyond software and digital workflows into the physical world.

Sector outlook

  • Semiconductors and hardware: Demand continues to exceed supply

We expect AI infrastructure demand to continue to outpace supply through 2027. Demand remains strong with hyperscalers and other customers providing suppliers with orders stretching through 2027 and beyond. 

Optical components, semiconductors, and power for data centres will continue to be key constraints on the supply side. Leadership will continue to diversify as investors look for the next area of investment to support training and inferencing applications, especially as model innovations open new opportunities.   

One key risk to monitor is access to capital at an economic cost, especially in a rising interest rate environment. It is unreasonable to expect the buildout to proceed in a linear fashion, but it is hard to know the timing of any digestion period or interruption in the AI infrastructure race. 

Given the magnitude of outperformance of technology infrastructure names particularly during the first half of 2026, it would not be surprising to see continued volatility through 2027 as investors debate the relative merits of software versus semiconductors.

  • Cloud service providers: Return on investment comes into focus.

2027 could be the year that cloud service providers prove the return on investment for their unprecedented AI infrastructure spending.

Cloud service providers are uniquely positioned to monetise AI given their unmatched data centre footprints, access to graphics processing units, rapidly expanding custom silicon programmes, and highly profitable core businesses which support funding needs.

Investors should become more constructive as cloud revenue growth accelerates, capex growth moderates, and cloud margins remain healthy. Together, these factors could provide compelling evidence that AI infrastructure can generate attractive long-term returns rather than simply support an arms race.

  • Software: Careful navigation required

The software sector will require increasingly careful navigation as investors distinguish between AI beneficiaries and companies vulnerable to AI disruption.

Broad-based scepticism toward application software should give way to a more nuanced assessment of which companies can leverage AI to expand their total addressable markets and strengthen their competitive positions.

While AI monetisation is likely to accelerate across the industry, it will still represent a relatively small portion of revenue for most software companies in 2027.

We remain constructive on cybersecurity and infrastructure software, where AI is driving both increased demand and new product opportunities, though we are mindful of premium valuations given strong performance in 2026.

  • The first wave of frontier AI IPOs arrives

The potential public listings of leading frontier AI labs in the year ahead would mark a significant milestone. The initial public offerings would likely represent some of the most notable technology debuts since the emergence of the cloud service providers.

Beyond expanding the investable universe, these listings could provide investors with direct pure play exposure to the development and monetisation of frontier AI models, potentially attracting significant capital flows and reshaping technology indices.

More broadly, due to public market scrutiny, these companies will be under pressure to demonstrate sustainable revenue growth, scalability, margins, and returns on capital.

Valuations bolster environment

Valuations across much of the technology sector remain supportive, providing a constructive backdrop for investors.

While select AI beneficiaries continue to command premium multiples, many software, semiconductor, and hardware names trade at levels that appear reasonable relative to their growth and earnings outlooks.

The average forward price-to-earnings ratio for the semiconductor industry has declined to below 19 times (as of mid-October 2026) compared to the five-year average closer to 25, and the software sector’s enterprise value-to-sales ratio is moderately below its long-term average.2 

  • Semiconductor stock valuations look reasonable given earnings momentum.  In the semiconductor industry group, valuations are compelling, but for companies in commodity markets for standardised components like memory chips, investors are hesitant to put high multiples on what could prove to be peak earnings expectations
  • Valuations remain compelling for select application software companies that can monetise AI and defend their competitive positions against disruption.Outside of application software,widely perceived AI beneficiaries including certain data management and cybersecurity companies now command premium valuation multiples. Meanwhile, the cloud service providers appear undervalued, particularly if further evidence emerges that AI investments are generating attractive returns
  • Select hardware stocks are undervalued. Optical and other data centre infrastructure stocks have pulled back on concerns about the sustainability of supply shortages and of demand, and perhaps in relation to the prospects of a Democratic sweep in the US (which we believe would not in and of itself have a major impact on data centre approvals); select stocks look compelling

The long-term opportunity created by AI remains significant, but investors should remain mindful of several risks that could slow adoption or pressure valuations.While these risks are worth monitoring, the balance of evidence continues to support a constructive outlook for the sector. Key risks include, but are not limited to:

  • High interest rates could result in valuation pressure for long duration assets and could raise the cost of capital for financing AI infrastructure
  • The geopolitical environment is uncertain and makes it more likely that there will be competing AI ecosystems, inefficient supply chains, and physical risks to AI infrastructure
  • AI safety concerns could slow development, particularly if there is a cyber or biological attack by a bad actor, and/ or if there are more breaches by AI agents going rogue
  • Financing challenges could slow AI investment, especially as a higher portion of capex is being debt financed
  • Trends could give rise to the formation of an AI bubble, including circular relationships, increasing use of debt financing and off-balance sheet vehicles, lax regulation, private market valuations moving into the public markets with IPOs, and rising interest rates (which could pop a bubble)

However, mitigating factors include investor scepticism, encouraging AI adoption trends in consumer and enterprise applications which demonstrate paths to positive return on investment, and reasonable valuations that are not yet at dotcom bubble levels.

The next stage

The next phase of the AI cycle is likely to be defined less by technological breakthroughs and more by execution, adoption, and monetisation.

Companies that can translate AI investment into sustainable revenue growth and attractive returns on capital should emerge as the long-term winners.

While periodic volatility is inevitable, we remain constructive on the technology sector and the opportunities created by the ongoing AI transformation.

[1] The choices we make about AI now are critical | Bill Gates / Dario Amodei — We Must Pace the Frontier / https://mustafa-suleyman.ai/a-warning-about-model-welfare

[2] Source: BNP Paribas Asset Management, FactSet.

Important information

Please note that articles may contain technical language. For this reason, they may not be suitable for readers without professional investment experience. Any views expressed here are those of the author as of the date of publication, are based on available information, and are subject to change without notice. Individual portfolio management teams may hold different views and may take different investment decisions for different clients. This document does not constitute investment advice. The value of investments and the income they generate may go down as well as up and it is possible that investors will not recover their initial outlay. Past performance is no guarantee for future returns. Investing in emerging markets, or specialised or restricted sectors is likely to be subject to a higher-than-average volatility due to a high degree of concentration, greater uncertainty because less information is available, there is less liquidity or due to greater sensitivity to changes in market conditions (social, political and economic conditions). Some emerging markets offer less security than the majority of international developed markets. For this reason, services for portfolio transactions, liquidation and conservation on behalf of funds invested in emerging markets may carry greater risk.

Back to Top