Tech stocks and “dip” don’t go together in this environment. However, stocks like Booz Allen Hamilton (NYSE:BAH), Reddit (NYSE:RDDT), and Arm Holdings (NASDAQ:ARM) are indeed deeply discounted at the moment, and it’s worth taking a deeper look at them.
Not every tech stock fits neatly into what the market thinks is ideal. This means if a tech company didn’t push AI hard enough in their earnings calls or had a bad quarter, that’s reason enough for Wall Street to shave off a large chunk of that stock’s premium. Buying these tech stocks at these lows is a good idea. When solid companies take a bruising, it usually turns out to be superficial, and dip buyers end up winning big.
I’d still be careful. You never know if a stock is going to get stuck like Zoom (NASDAQ:ZM) or PayPal (NASDAQ:PYPL). The best way to avoid falling into that pitfall is to buy into businesses with a high or accelerating growth rate.
Here are the AI tech stocks that are unlikely to disappoint.
Booz Allen Hamilton (BAH)
Booz Allen Hamilton is Palantir’s (NASDAQ:PLTR | PLTR Price Prediction) struggling twin, but one I don’t think will struggle for long. Of course, Palantir is more software-oriented, though both companies still do similar work. Booz Allen is a tech consulting firm that derives almost all of its revenue from the government. For most of its history, this was a major plus point and a source of stability that investors rewarded the company for.
That turned into a liability after November 2024 due to unclear and rapidly changing statements about the government’s priorities. The government introduced aggressive DOGE cuts and spoke about austerity measures, before suddenly being in favor of over $1 trillion in defense spending. Earlier this year, the U.S. Department of the Treasury cancelled 31 contracts with Booz Allen Hamilton, valued at roughly $21 million in total obligations and $4.8 million annually.
Despite these setbacks, I see solid potential ahead as the government is no longer cutting and increasingly needs more guidance from firms like Booz Allen in order to integrate AI. Both revenue and EPS are expected to rebound starting next year.
BAH stock is down 59% from its high, and I expect a full recovery.
Reddit (RDDT)
Reddit is down nearly 50% from its peak, and it’s both a recovery and a growth play at the same time. The company’s market cap is just $26 billion, and I think that undersells how much sway this platform has to a great margin.
If you look at the most visited websites in the U.S., Reddit ranks fifth. This platform gets more traffic than even Instagram, ChatGPT, or X. Better yet, Reddit is increasingly popular among younger individuals, and I expect it to eventually become the third-most visited website as the Facebook demographic ages out.
On top of that, Reddit’s platform is a treasure trove for AI due to the amount of conversational data it has. It remains under-monetized as it doesn’t serve many ads and doesn’t push a subscription as hard. I believe that will change in the future as more platforms block ad blockers and introduce subscriptions. Reddit’s U.S. average revenue per user (ARPU) grew 126% to $10.79 in Q4 2025. I expect plenty more room for growth, as Facebook drew in nearly $60 per user.
I’d buy the dip.
Arm Holdings (ARM)
Arm Holdings designs and licenses high-performance CPU chips. It does not manufacture any chips, but its designs are very popular with portable devices like phones and laptops. These chips are designed to be highly energy-efficient while being performant, and they might be the perfect crossover for what AI needs today.
Companies just need to use the ARM architecture to benefit the company. Even Apple’s (NASDAQ:AAPL) M-series chips are based on the ARM architecture, and Apple pays royalties for every chip.
Battery technology has lagged behind basically everything else. Your phone or your laptop has better memory, GPU, CPU, but probably not a better battery compared to a decade ago. The longer battery life you do get comes mostly from more efficient chips and software. This is where Arm comes in, as they specifically make low-power chips, and ones that can even handle AI models that run on the phone locally without internet.
Per CEO Rene Haas, Arm’s platform is roughly 50% more efficient than competitive solutions. This is an enormous advantage when hyperscalers are racing to stand up AI data centers, but cannot get enough electricity to the building. In that same quarter, Arm’s data center royalties doubled year-over-year, as hyperscaler companies deployed Arm-based chips.
Q3 FY 2026 royalties reached $737 million, up 27%. It now constitutes a majority of the $1.24 billion in total sales.
I see triple-digit upside as analysts see revenue growth holding above 20% for the coming years.