⚖️ Read AI oversight order that President Donald Trump said ‘didn't like certain aspects of it’ ↗
A draft AI executive order surfaced after the planned signing was abruptly shelved. It would have allowed federal agencies to review powerful frontier models before release, mainly for cybersecurity and national security risks.
The notable bit was its emphasis on the process being voluntary, not a licensing regime. Still, the politics turned sticky fast - innovation, China, safety, all crammed into one policy suitcase that wouldn’t quite shut.
🕵️ OpenAI launches free AI image verification tool amid deepfake surge ↗
OpenAI rolled out a public image-checking tool designed to help people identify whether an image came from AI systems. It leans on invisible watermarking and metadata signals, including SynthID and C2PA.
This is very much a deepfake-era move. Handy, though not magic - metadata can be stripped, screenshots can muddy the water, and bad actors do enjoy being bad actors.
🌍 Scotland’s ‘green datacentres’ policy ignores emissions impact of AI, analysis shows ↗
Scotland’s push for “green datacentres” came under pressure after analysis argued the policy doesn’t properly account for AI’s huge energy appetite. More than a dozen projects are reportedly moving through planning.
🧼 ‘AI washing’: firms are scrambling to rebrand themselves as tech-focused ↗
Companies are apparently bending themselves into pretzels to sound like AI firms, even when they’re mostly selling ordinary automation. PR people are getting dragged into the theatre too, which sounds exhausting.
The term “AI-powered” is doing a lot of heavy lifting. Sometimes it means generative systems, sometimes it means a scanner with better software, and sometimes… it’s just glitter on a toaster.
🎬 ‘We’re expanding the cinematic toolbox’: AI fault lines on show at Cannes ↗
AI became one of the loudest arguments around Cannes, with film-makers split between “new creative tool” and “please, absolutely not.” Darren Aronofsky defended AI as additive, while other voices pushed back hard.
The interesting part is how practical the debate has become. It’s not just robot screenwriters anymore - it’s AI visuals, digital doubles, production shortcuts, and the slippery question of what still feels human on screen
FAQ
What was in the stalled AI oversight order linked to Donald Trump?
The draft AI executive order would have allowed federal agencies to review powerful frontier AI models before public release. Its main focus was cybersecurity and national security risk. The article notes that the process was framed as voluntary, rather than as a formal licensing regime, but the politics around innovation, China, and safety made the proposal difficult.
Why does voluntary AI oversight still cause political debate?
Voluntary AI oversight can still feel sensitive because it sits between safety review and government control. Supporters may see it as a way to catch serious risks before release. Critics may worry that it slows innovation or gives rivals an advantage. In this case, the article suggests those tensions helped make the order politically awkward.
How does OpenAI’s AI image verification tool work?
OpenAI’s public image-checking tool is designed to help people identify whether an image may have come from AI systems. It uses signals such as invisible watermarking and metadata, including SynthID and C2PA. The article presents it as helpful but limited, because metadata can be removed, screenshots can reduce clarity, and bad actors may try to avoid detection.
Why are AI deepfakes hard to verify reliably?
AI deepfakes are hard to verify because the clues used for detection are not always preserved. Metadata can be stripped when files are edited, reposted, compressed, or screenshotted. Watermarks can help, but they are not a complete guarantee. A common approach is to combine technical checks with source verification and basic media-literacy habits.
What is the concern about Scotland’s green datacentres and AI?
The concern is that Scotland’s “green datacentres” policy may not fully account for AI’s large energy demand. The article says more than a dozen projects are reportedly moving through planning. Critics worry that climate-friendly branding could make projects sound cleaner than they are, especially if they rely on major power use or backup diesel infrastructure.
What does AI washing mean in business?
AI washing means companies describe products or services as AI-focused even when the technology may be ordinary automation or only lightly AI-related. The article says firms are rebranding themselves to sound more tech-driven. Terms like “AI-powered” can be vague, so buyers and readers should ask what the system does in practice and where AI is meaningfully involved.