The biggest focus in the tech industry in 2026 will be where the balance settles between aggressive AI investment and growing talk of a bubble. For now, uncertainty is extremely high. Talk of a bubble and aggressive investment could continue to coexist in a standoff, or one side could abruptly lose momentum.

In the smartphone market, key points to watch have emerged over whether Apple can reverse sentiment with AI and whether Samsung Electronics will join the foldable phone race it moved into first. Stablecoins, which became an issue across the financial industry in 2025, are also expected to retain their box-office power in 2026. The focus is on how far they can turn potential as a payment method into reality beyond cryptocurrency trading.

Will the AI bubble burst?

This year, voices warning about an AI bubble are likely to continue. Calls to be wary remain, citing the lack of results commensurate with excessive investment. Even so, many still appear to view the likelihood of a bubble materialising as low. Fortunes may diverge among individual companies, but the prevailing view is that a scenario in which the entire AI industry faces a crisis would be an overreaction.

Venture capitalist Ben Horowitz, who leads prominent Silicon Valley venture firm Andreessen Horowitz, said: “A bubble forms when everyone believes prices cannot fall. The clearest evidence we are not actually in a bubble is that everyone is talking about a bubble.”

Casey Newton, founder and editor of Platformer, wrote recently in an outlook for 2026 that talk of an AI bubble will continue in 2026. He predicted that some companies, especially startups, could collapse in the process, but the bubble will not burst.

Can Google Gemini overtake OpenAI ChatGPT?

In terms of the competitive landscape in the AI industry, interest is high in whether OpenAI will keep its edge or whether Google, which has been biding its time, can pull off a comeback.

After OpenAI released ChatGPT in late 2022, Google appeared to fall behind in the AI race. It then began to narrow the gap in earnest from the launch of the Gemini 3 model in the second half of 2025. Attention is now on how much Google can accelerate its pursuit this year.

Stephanie Palazzolo of The Information drew attention by predicting that Gemini will even surpass ChatGPT this year in weekly active users. She thinks a turnaround against ChatGPT would not be difficult if Google pushes AI Mode and Gemini more aggressively to Google Search users.

Can Apple restore its reputation with Siri?

Apple’s role in this year’s AI race is also of interest. Apple has been relatively sidelined by the AI boom, but in 2026 it will seek to restore its reputation with a next-generation Siri armed with generative AI.

An upgrade to a generative AI-based Siri was originally planned for last year, but it was pushed back to this year after failing to meet internal standards.

Apple is said to be developing the next-generation Siri based on a new architecture. With ChatGPT and Gemini also offering voice functions, attention is on whether Apple can deliver a usable result.

How far will AI hacking evolve?

Signals are being detected one after another that AI hacking tools are evolving into increasingly powerful forms. In the past there was plenty of talk that AI hacking tools would not be easy to use in real-world attacks, but the recent situation appears to be changing.

The Wall Street Journal reported that even hackers with limited technical knowledge can use AI models to generate convincing deepfakes or create websites and emails that steal passwords and two-factor authentication codes. As AI assistants are built into web browsers, malicious actors have also become able to hide malicious commands in web pages designed to trick AI assistants.

Trevor Hilligoss, a former FBI cyber task force officer and now senior vice president at security company SpyCloud Labs, said: “AI makes it easier for more people to access better malware, and that is truly terrible.”

Google has also found bot-based software that can obfuscate its own code to evade detection and generate new malicious functions, the WSJ reported. Newton predicted that in 2026, state-level attackers and cybercriminal groups will weaponise LLMs to inflict serious damage and provide the concrete evidence that AI safety advocates have been seeking to press for stronger regulation of AI development.

How will AI agents reshape the B2B SaaS ecosystem?

AI agents that use LLMs to automate specific tasks without human intervention, rather than simple chatbots, are also a major tech keyword for 2026. If AI agents prove they work in real-world use, bubble fears could subside and the view of AI as the dominant trend could gain more momentum. If the opposite occurs, talk of an AI bubble is likely to remain an issue next year as well.

AI agents are also a major variable for enterprise software business models. Some forecasts say software in systems of record that store and manage data, such as ERP, CRM, HR and ITSM, will undergo fundamental change because of AI agents. Many also predict that some subscription-based software-as-a-service products could see their position weakened.

Andreessen Horowitz, also known as a16z, cited one notable 2026 scenario as systems of record in enterprise software finally beginning to lose control. AI is narrowing the gap between intent and execution. The firm said models can now read, write and reason over operational data directly, meaning software such as ITSM and CRM will shift from passive databases to autonomous workflow engines.

Robotaxi expansion faces a test

In 2025, self-driving cars made major progress in terms of popularisation. Waymo expanded the areas where it operates robotaxi services, and Amazon also joined the race. Amazon has been testing a robotaxi service in Las Vegas since September and plans to expand to Miami and Austin, Texas.

But it is difficult to conclude that robotaxi services are smoothly heading toward mass adoption based on the current situation alone. Semafor predicted autonomous driving will face major change this year, and that could be positive or negative. It said if self-driving cars become common, accidents involving not only pets but also humans will be more likely, and it is unclear whether self-driving cars are established enough to withstand such incidents. Semafor reported that one accident derailed Uber’s ambitions for autonomous driving and that Cruise is barely hanging on because of accidents.

Foldable iPhone finally becomes reality

If current expectations hold, Apple is likely to join the foldable smartphone race pioneered by Samsung Electronics this year. Foreign media including Bloomberg News and The Information have reported Apple is expected to unveil a foldable iPhone this fall as part of the iPhone 18 lineup.

The foldable iPhone is expected to take a book-style form factor, include multiple cameras and use Touch ID instead of Face ID. Unlike foldable phones offered by Samsung and Google, Apple’s foldable model is also reported to have a screen ratio similar to the largest iPad when viewed in landscape mode.

The foldable phone is also drawing attention because it is expected to be the most complex iPhone Apple has developed. Apple is said to have delayed the launch of a foldable iPad that was due around a similar time early this year and reassigned related engineers to the foldable iPhone project to meet the schedule. The Information reported Apple is also likely to apply a premium price to the foldable device, as Samsung and Google have priced their foldable models at a premium compared with other models.

Stablecoins move beyond crypto toward payment infrastructure

As enthusiasm for stablecoins grew last year, moves by related industries to deploy stablecoins as payment infrastructure accelerated. For now, optimistic and sceptical outlooks coexist on the use of stablecoins as a payment method.

A report JPMorgan wrote last year said most stablecoin demand still comes from within the cryptocurrency ecosystem. It analysed that so-called “crypto-native” activity, including trading, DeFi collateral and idle funds held by crypto firms, accounts for 88 percent of total demand, while real-world payments are only 6 percent.

Even so, the industry is moving faster to use stablecoins for cross-border remittances and settlement, and attention is on whether that will lead to tangible results. Hasu Qureshi, managing partner at crypto venture capital firm Dragonfly, sees new payment rails accelerating stablecoin adoption. He predicted these channels will play a key role in making stablecoins a routine means of payment, particularly in emerging markets.