The AI disruption in software has been one of the major anxiety-inducing themes of 2026. And while the negative momentum in some of the software plays has since slowed in the month of March, the rest of the market starts gravitating lower due to higher oil prices (an oil shock of sorts) and the war in Iran.
In any case, the software slide may or may not be over with quite yet. And while there was a bit of a rally in some of the oversold names, including Thomson Reuters (TSX: TRI), I’m just not so sure it’s safe to go bottom-fishing. Not while the gains from late-February are in the process of evapourating. While Thomson Reuters is doing a great job of pivoting towards AI, it’s just really difficult to tell whether the rise of agentic AI, specialized AI tools, and vibe-coding is disruptive enough of a headwind to get in the way of a sustained comeback in the exposed software plays.
Source: Getty Images
Thomson Reuters has AI and agents of its own. But is that enough to justify buying the dip?
While Thomson Reuters’s own CoCounsel and agents are powerful and a great, convenient way for users to get more work done faster, I think that it might not matter how impressive these tools are if the playing field has been evened out and such tools are in the hands of more users who might no longer be interested in paying for a seat.
Of course, time will tell if the AI disruption is going to take away Thomson Reuters’s previously wide economic moat. Because it’s hard to tell, I think it might be wise to be cautious on the name, especially as the negative momentum accelerates again and AI tools become even more capable.
It’s tough to tell how this ends for the software companies. While they’re probably not going to zero come the next big Anthropic AI tool release, I think that the fact that such AI innovators have set their sights on legal software may act as a bit of a warning shot for investors looking to back up the truck on the dip. Thomson Reuters is kicking off a huge buyback program, with a plan to repurchase US$600 million worth of shares through August.
That’s a massive vote of confidence. But, then again, the firm isn’t the only one that’s viewing the latest crash in the stock as a buying opportunity. Personally, I would not follow the firm in, even if they’re confident that their own AI tools can keep the economic moat intact and offset any potential AI disruptive pressures that have arisen in recent months.
Shares are cheap. Or are they?
While there might be colossal rewards by loading up on weakness, shares of TRI are on the cheaper end of the historical range at 27.4 times trailing price to earnings (P/E), investors who aren’t sure must be humble and should acknowledge the potential risks of being wrong.
What happens if AI’s disruptive impact really does justify the valuation reset in software names? Then, perhaps $125 per share might be more of a fair value and less of a bargain. It’s hard to tell. And for that reason, I’m going to sit on the sidelines, even if it means missing out on a historic discount and a potential V-shaped bounce.
