THE RUNDOWN

Elon Musk's AI company xAI has filed a lawsuit against a South Carolina man who allegedly used the Grok chatbot to generate child sexual abuse material (CSAM) deepfakes. It's one of the first high-profile cases where an AI company has gone after a user — rather than the other way around — for misusing its tools. The case raises serious questions about what guardrails AI companies are legally and morally responsible for building into their products from day one.

Why it matters for Ghana: As AI tools become more accessible across Africa, this case sets an important precedent for how AI companies will respond when their platforms are weaponised — and what liability looks like in the age of generative AI.

Microsoft is reportedly coaching its salespeople to position its in-house AI models as cheaper and more efficient alternatives to OpenAI's GPT models and Anthropic's Claude. This is striking because Microsoft is still one of OpenAI's biggest investors and partners. It signals that the corporate bromance is quietly souring as Microsoft pushes to own more of the AI revenue stack itself.

Why it matters for Ghana: For African businesses and developers evaluating AI tools, a more competitive Microsoft offering could mean better pricing and more choice — worth watching closely.

SpaceX shares have been sliding steadily from their post-IPO highs, settling around the $135 mark ahead of a planned Starship launch. Markets seem to be reality-checking the ambitious promises Elon Musk made before and after the listing. It's a reminder that even the most hyped tech companies eventually have to face sober investor expectations.

On top of everything else, Elon Musk has quietly purchased a gas turbine company, and the most likely reason is power — literally. xAI's data centres are notoriously energy-hungry, and owning a direct turbine power source would give SpaceX and xAI more control over their electricity supply. It's vertical integration taken to a dramatic extreme.

Why it matters for Ghana: Africa's AI ambitions are consistently constrained by unreliable power. Watching how hyperscalers solve energy problems at scale could offer real lessons for local data centre development.

While fighting a legal battle with Apple over alleged hardware trade secret theft, OpenAI has released a light-up physical keyboard built in collaboration with Work Louder, designed specifically for controlling its Codex AI coding agent. It's an interesting move into hardware territory — and the irony of launching a physical product mid-lawsuit with Apple isn't lost on anyone. At $230, it's clearly aimed at developers and power users rather than the mass market.

Samsung says it's internally investigating reports that some Galaxy S26 Ultra displays are developing a reddish discoloration. The company hasn't confirmed why it's happening yet, which is not a great look for its top-tier flagship. If you're considering picking up an S26 Ultra, it's probably worth waiting to see what Samsung says officially before spending that kind of money.

Why it matters for Ghana: Samsung is among the most popular premium Android brands across West Africa. Defects like this affect resale value and long-term trust — and local buyers deserve to know before purchasing.

In a win for the right-to-repair movement, Valve has confirmed that iFixit will continue selling Steam Deck replacement batteries after briefly appearing to wind down the partnership. Valve has built a genuinely repair-friendly reputation, so the initial news of discontinuing the battery was jarring. It seems the backlash was enough to reverse the decision — which is exactly how consumer pressure is supposed to work.

Microsoft has patched a security flaw in the decades-old real-time strategy game Age of Empires II that could have let attackers take over a victim's computer simply by sending a malicious game invite. It's a reminder that old software carries old risks — and that gaming platforms are increasingly attractive targets for security threats. The patch is out now, so if you're still playing AoE II (no judgement), make sure you've updated.

With Christopher Nolan's Odyssey adaptation hitting cinemas this weekend, critics are also pointing to a wave of cheaply produced, AI-generated movies flooding streaming platforms — essentially the 2026 version of direct-to-video cash grabs. Studios are using generative AI to churn out low-effort content designed to ride trending search terms rather than tell actual stories. It's a genuine concern for the film industry and for audiences who increasingly can't tell what's real craft and what's automated filler.

A federal judge has ruled that the Trump administration cannot revoke visas or deport researchers simply for working in content moderation or disinformation research. The ruling was praised by academic and civil society groups who said the administration had been using immigration policy to silence critics and researchers who study online misinformation. It's a significant moment for academic freedom in the US — and has implications for how tech policy research is protected globally.

That's your July 16 rundown. Between AI lawsuits, a stock market reality check on SpaceX, and a screen defect Samsung can't yet explain, it's been a day that reminds you tech moves fast — but consequences have a way of catching up.