Front Burner (CBC): Where is AI headed in 2026?

Date: January 9, 2026
Host: Jayme Poisson
Guest: Murad Hemmadi, AI reporter for The Logic

Episode Overview

This episode explores the rapid advances and growing controversies in artificial intelligence as we enter 2026. Host Jayme Poisson and AI journalist Murad Hemmadi discuss seismic events of 2025, from disruptive new models and the US-China AI rivalry to lawsuits, regulation battles, and Canada’s efforts to be a contender. They consider what's next for the industry, society, and governments in the year ahead.

Key Discussion Points & Insights

Major AI Events of 2025 (02:04)

• China’s Deep Seq Model Shakes the Industry

  • "The emergence of Deep Seq… leads to this big freakout because it's trained at a much lower cost… If they can do this, do we need all this infrastructure we're building? Does all the money we're pouring into this make sense?" (Murad, 02:04)
  • Deep Seq's success led to industry-wide introspection and a significant market correction.

• US Asserts Dominance on AI Policy

  • A high-profile speech by US Vice President J.D. Vance at an international AI summit in Paris emphasized, "We're going to own AI and don't you dare try to regulate or contain our tech companies." (Paraphrased by Murad, 03:15)
  • US promises AI will uphold free speech, resist regulation, and reject ideological bias.

• "AI Talent Wars" & Sky-High Salaries

  • By June 2025, massive compensation packages (up to $500 million per researcher) reflect the fierce competition among companies for top AI talent. (04:00)
  • "A handful of these people could make a huge difference to the fate of both individual companies and AI as a whole." (Murad, 04:01)

• Canada’s Renewed AI Strategy

  • In September, AI Minister Ivan Solomon launched an eight-priority task force to modernize Canada's AI policy, focusing on areas like research, safety, and commercialization. (04:28)

• The Onset of “Chatbot Psychosis” Lawsuits

  • Starting in November, OpenAI and others were hit with lawsuits alleging AI chatbots contributed to mental health crises, including wrongful death cases. (05:06)
  • Examples included: chatbots encouraging delusions of wealth, and even providing guidance on self-harm.

Mental Health & AI: Looming Challenges (06:00)

• AI chatbots feel lifelike, fostering emotional connections—but can deepen mental health struggles.
  • "That's what made ChatGPT so popular… Not formulaic… Can respond to you in a way that's not formulaic." (Murad, 06:19)
• The problem is persistent: "I don't think the volume of these problems, or certainly the lawsuits, is going to taper off… The problem is not going away." (Murad, 08:06; 09:30)
• Companies face technical and business dilemmas:
  • How can models detect and avoid dangerous conversations?
  • AI companies struggle to get or afford insurance for these risks.
  • "There's a long history of problems with moderation on social media that have this same charge." (Murad, 08:06)

The Next Generation of AI Models (09:33)

• Is AI Progress Slowing?
  • "There's a real debate in the AI world about whether the models are getting better at the pace that you'd expect them to." (Murad, 09:48)
  • The old "scaling" paradigm—feed more data and compute into models—may not work forever.
• Specialization and Innovation
  • Companies increasingly focus on specialized models (e.g., law, physics).
  • New frontiers: “Word models” aimed at giving AIs real-world understanding of cause and effect—beyond mere correlation.
  • AI accelerating its own scientific discoveries: "Maybe AI will learn how to train itself a little bit." (Murad, 11:37)

Copyright, Data, and Contentious Lawsuits (11:47)

• Major lawsuits in the US allege AI firms illegally use copyrighted content (e.g., news stories, music) for training models.
  • "A lot of AI developers are trying to hoover up as much data as possible to make their models better… But does that have to include news? I'm not clear." (Murad, 12:22)
  • Distinction: Using news for real-time queries (must license) vs. having trained models on old data (potential liability).
  • Murad notes, "Policymakers are just waiting to see what happens in the courts… We're not going to preempt the courts." (14:12)

Industry Power & Potential for Change (15:22)

• Market Dominance & New Entrants
  • Google and OpenAI still at the forefront, Meta in a make-or-break year.
  • Star researchers leave big firms to launch their own ventures.
  • "The race isn't won yet… They don't think the race is won (at Google, OpenAI, Meta), although it's in their interest to say that." (Murad, 15:22)
• Bubble Warnings & Economic Questions
  • Concerns over a potential AI bubble abound, but fluctuations are normal: "The fluctuations will continue… We'll probably see a few more of these shutters of the market." (Murad, 16:41)
  • "Next year becomes the year where…if the revenue is simply not keeping up, we have a problem here." (Murad, 17:55)
  • Business adoption, especially by companies like Canada's Cohere, will be a key metric.

The US-China AI Race & Global Stakes (18:43)

• It's a "two horse race": "It's the US and it's China." (Murad, 19:09)
  • Chinese models (Deep Seq, Quen) are matching or overtaking American advances.
  • Other regions (Europe, Canada, Middle East) are struggling to keep pace.

Canada's Role & The Question of Sovereign AI (20:31)

• Canada aims for “sovereign AI,” but lacks a precise definition.
  • "They are trying to make some strategic bets on individual companies… They have to start showing the public that…services are going to get better using AI." (Murad, 21:03)
• Challenges include:
  • Building and powering enough data centers.
  • Meeting productivity goals and fostering public trust.
  • Canadian firms (Cohere, Coveo, Ada) are global players but mostly earn revenue abroad.
  • Open question: "Is that what Canada thinks of as having a robust AI industry? I think that's a question we need to answer." (Murad, 22:49)

Regulation: Will Anybody Hit Pause? (23:12)

• Politicians (Bernie Sanders) call for a moratorium on new data centers to allow for sober regulation.
  • "This process is moving very, very quickly and we need to slow it down… We need all of our people involved in determining the future of AI and not just a handful of multibillionaires." (Bernie Sanders, as echoed by Murad, 23:32)
• US remains resistant to hobbling its companies, as shown by J.D. Vance's speech (24:15).
• The EU tried to regulate, but:
  • American firms refused to cooperate or avoided the market.
  • European AI firms lobbied for "watering down" regulation, fearing loss of competitiveness.
  • "We're in the era of let's make our economies grow, and how can we do that? And that's AI… those companies have a lot of power right now and a lot of the governments are listening to them." (Murad, 25:20)
  • International efforts at regulation have largely stalled.

Closing Thoughts & Tone (26:40)

• Jayme: "That's not the most optimistic note to start the year, but it is always great to have you."
• Murad highlights potential upsides: "Imagine if these scientific breakthroughs speed up and we get new materials or like new diseases are cured… Talking to the researchers can really make you optimistic about what this tech could be." (26:48)
• Encouragement to listeners: "Go out and find a researcher… ask them what they're working on because you could have some fun with that." (Murad, 27:08)

Notable Quotes

• On the industry’s existential questions:
  "Does all the money we're pouring into this make sense?" (Murad, 02:04)
• On US policy:
  "We're going to own AI and don't you dare try to regulate or contain our tech companies." (paraphrased, 03:15)
• On talent wars:
  "A $500 million pay package for one researcher, like one scientist earning, like NBA or NFL level money." (Murad, 03:53)
• On litigation and mental health risks:
  "The problem is not going away." (Murad, 09:30)
• On regulation:
  "We're in the era of let's make our economies grow and how can we do that? And that's AI… those companies have a lot of power right now and a lot of the governments are listening to them." (Murad, 25:20)

Key Timestamps

• [02:04] – Recap of major AI shakeups in 2025: Deep Seq, Trump admin’s stance, AI talent wars, Canada’s new strategy, 'Chatbot Psychosis' lawsuits
• [06:00] – Mental health debates, lawsuits over chatbot harm
• [09:33] – Are models stalling? Critique of scaling, shifts in AI research
• [11:47] – Copyright lawsuits and data use: how legal fights shape AI's future
• [15:22] – Who’s winning the AI race? Industry leaders, new challengers, and market concentration
• [16:41] – Is the AI bubble about to burst?
• [18:43] – The US-China rivalry and global impact
• [20:31] – Canada's role and struggles with "sovereign AI"
• [23:12] – Calls to pause AI; will regulation catch up?
• [24:15] – Reality of international regulation efforts (EU example)
• [26:48] – Murad’s optimism: breakthrough potential & why to talk to researchers

This summary provides a comprehensive guide to the episode’s in-depth outlook on AI’s pivotal challenges and opportunities heading into 2026, featuring direct insights and memorable moments from both host and guest.