AI predictions for 2026

“In the midterms, AI will be an election issue for first time.”

“In 2026, look for a nation-state or group of cybercriminals to weaponize an LLM to cause significant damage, and give AI safety advocates the concrete evidence they’ve been looking for to push for stronger regulations on AI development.”

“One of the large global pharma companies will acquire one of the leading protein AI startups.”

“Sam Altman will step aside as CEO of OpenAI.”

“OpenAI, too, will eventually become a publicly traded company. But not in 2026.”

“In 2026, Anthropic will debut on public markets. It will be one of the biggest and most highly anticipated IPOs of all-time.”

“Models like Flux, Ideogram, Leonardo, Midjourney, and Reve will likely be acquired by multimodal AI labs such as Google, Meta, OpenAI, and xAI.”

“For some reason, the big AI labs haven’t released music models yet, but this is likely to happen in 2026.”

“By the end of 2026, AI will likely perform tasks that take a human 39 hours which is more than a full work week when you account for meetings and other non-task activities.”

“Specifically, I predict that news mentions of “AI” and “suicide” in media databases will be at least three times higher in 2026 than in 2025.”

“More precisely: there will be no unusual physical or economic catastrophe (dramatically larger than past incidents of a similar nature) in which AI plays a crucial enabling role.”

“Forecasters estimate that there is an aggregate 9.6% (5% to 15%) probability that a Google DeepMind model will be among the top 3 models at the end of 2026, according to Epoch’s capability index.”

“By the end of 2027 , AI will complete tasks that take a human two months, doing so in perhaps an hour while consuming roughly $100 worth of compute.”

“If there exist AI systems that can perform 8 of the 10 tasks below by the end of 2027, as determined by our panel of judges, Gary will donate $2,000 to a charity of Miles’ choice; if AI can do fewer than 8, Miles will donate $20,000 to a charity of Gary’s choice.”

“We won’t get to AGI in 2026 (or 7).”

“My assessment is that there’s roughly a 30% chance of a significant AI-driven market correction with at least a >20% drawdown in AI-heavy stocks, sometime within the next three years.”

“Our forecasters think there’s a 5.2% (2% to 10%) chance that the US Senate will pause frontier AI development in the US before 2029, absent a treaty with China.”

“If the AI gets an A on at least 5 of out 6 of those exams using same grading scale as his students, then Bryan owes Matthew $500. Otherwise, Matthew owes Bryan $500. For the purpose of this bet, an A- counts as an A.”

“By the end of 2030 (when superintelligence is expected), AI will perform tasks that take humans approximately 100 years to complete, doing so in less than a day.”