NEWS YOU CAN’T MISS

Your CEO Might Not Be Real: The $25M AI Scam

AI is fueling a new wave of scams in which criminals use deepfake voices and videos to impersonate CEOs and executives. These fake calls and messages look and sound so real that employees are being tricked into wiring millions of dollars or giving away sensitive data. 😱

With just minutes of public audio or video, scammers can train AI to mimic an executive’s voice, tone, and even expressions in real time.

Angus Loten reports in the Wall Street Journal that the US faced more than 105,000 deepfake-related attacks in 2024, about one every five minutes. Losses already topped $200 million in the first quarter of 2025.

🔎 Real-World Examples

• Arup (UK): An employee sent $25 million after joining a video meeting where the “executives” looked and sounded real but were actually AI fakes.

• Ferrari, Wiz, WPP: These global companies all reported fake “CEO messages,” showing even big, well-known brands can be targeted.

• YouTube: Scammers created a fake video of CEO Neal Mohan announcing new money-making rules. The clip tricked creators into clicking on fake websites that stole their passwords.

These scams succeed because they exploit trust and urgency - employees act fast when they think the boss is calling. The solution isn’t just tech: companies need sharper verification protocols, regular staff training, and strong deepfake detection to avoid becoming the next $25 million headline.

Sweden Strikes Back: A New AI Powerhouse

Five of Sweden’s most powerful players (AstraZeneca, Ericsson, Saab, SEB, and Wallenberg Investments) are joining forces to launch Sferical AI, a company that will build AI infrastructure that stays under European control.

Sferical isn’t trying to win the race for the biggest models, unlike OpenAI or Anthropic. Its focus is on building safe, transparent systems where data, models, and infrastructure stay within European borders, overseen by European governance, writes Ron Schmelzer.

Why this matters: it’s all part of a global trend called Sovereign AI. Countries are increasingly viewing advanced AI systems as critical infrastructure, (like energy grids or telecom networks). Relying solely on US tech giants or Chinese platforms is seen as a growing risk for governments, corporations, and national security.

Similar efforts are underway across Europe:

• Mistral AI in France has raised over €600M to build European-focused language models.

• In Germany, Aleph Alpha, backed by SAP and Bosch, is developing homegrown AI.

• And in India, BharatGPT is creating multilingual models tailored to local industries.

Sferical’s edge comes from direct access to high-value industry data: AstraZeneca (pharma), Ericsson (telecom), Saab (defense), SEB (finance), and Wallenberg Investments (capital).

In short, Sweden isn’t just building AI: it’s building sovereign AI muscle with industry power baked in. 😎

New Buzzword: What’s a “Clanker”?

A new AI insult is taking over TikTok: “clanker.”

Borrowed from Star Wars, the word has become the go-to way to mock everything from glitchy chatbots to delivery robots rolling down the street. Videos of teens yelling “clanker” at robots have pulled millions of views, and the slang is now spreading across Instagram and Reddit as well.

Gen Alpha (younger than 15yrs of age) is leading the charge.

For younger users, “clanker” works as both a meme and a subtle act of rebellion. It is a way to laugh at the flood of low-effort AI content filling feeds and to push back against automation they never asked for.

Even politics has picked it up. US Senator Ruben Gallego (D-AZ) used the term while promoting legislation to ensure people can always opt for a human on customer service calls, saying, “My new bill makes sure you get a real human on the phone … you don’t have to talk to a clanker if you don’t want to.”

So, where do you stand?