Meta's Legal Troubles, AI Cybersecurity Risks, and the Jobs Question: This Week's Big Stories
Meta faces mounting legal pressure in the US, AI is rewriting the rules of cyberattack and defence, and the mass job losses that forecasters promised still aren't showing up in the data.
Key points
- Meta is facing growing legal jeopardy in the United States, with cases that could reshape how the company operates.
- AI tools are lowering the cost of sophisticated cyberattacks, putting capabilities once reserved for governments within reach of criminal networks.
- Despite years of warnings, mass AI-driven job losses have not yet appeared in employment data.
- Tesla paid CEO Elon Musk 2.5 million times more than its average worker in 2025.
- Spotify plans to label AI artist profiles and remove them from recommendation queues.
What is happening to Meta legally?
Meta's legal exposure in the United States is growing across several fronts. The outcomes could affect how the company collects data and runs ads; at this scale, regulatory pressure tends to end in product changes, heavy fines, or both. We covered Meta's broader ambitions in our 14 August story on Zuckerberg's AI vision, which is now running directly into these legal headwinds.
Should you worry about AI and cybersecurity?
Yes, and the threat has a specific shape worth understanding. AI is now cheap enough that criminal networks can run attacks that once required state-level resources.
A convincing phishing email used to take a skilled writer hours. An AI model can generate thousands of personalised versions in minutes. Defenders are using the same tools to catch attacks faster, but the arms race is accelerating. AI Security has been our most-covered beat lately: 109 stories in the past 30 days, including 12 on phishing since 16 July.
The practical upshot: be more sceptical of urgent requests by email, text, or phone, even when they sound polished. AI makes the bait convincing, but the hook hasn't changed.
Where are all the lost jobs?
AI was supposed to hollow out entire job categories by now. It hasn't, at least not visibly.
Employment data shows no mass carnage. Economists point to a few reasons: AI tools are raising productivity without eliminating roles outright, adoption is slower than headlines suggest, and new work is appearing alongside automation. Disruption is arriving unevenly, and the people feeling it first rarely show up in aggregate statistics. Our 11 August story on Big Tech working hours found the opposite of the promised four-day week inside Google, OpenAI and Meta.
Survivorship bias matters too. Workers who got faster and better-paid with AI tools make good copy. The freelance translator who quietly lost clients doesn't.
Other stories worth noting
| Story | The short version |
|---|---|
| Spotify and AI artists | Labels AI artist profiles and removes them from recommendation queues; we've covered this since 14 July |
| Bumble rule change | Women-first messaging requirement dropped as app chases engagement |
| Roku AI channel | Early reviews describe it as low-quality, auto-generated video |
| Tesla pay gap | CEO Elon Musk earned 2.5 million times the average Tesla worker's pay in 2025 |
| US banning Chinese robots | Government moves to restrict certain robotics imports over security concerns |
One honest takeaway: The cybersecurity story is the one I'd watch most closely. Job disruption is real but slow; legal cases grind for years. AI-assisted scams are already in your inbox. The single most useful habit is pausing before you act on any urgent request. The bait's just gotten smarter.
