Adobe Puts Photoshop, 70+ Tools Inside ChatGPT
ALSO: Meta’s AI breaches a real company, OpenAI vs Apple
Welcome to Guru99 AI Report!
Top Story: Welcome back. A lot happened while you were busy: Adobe’s creative suite landed inside ChatGPT, AI quietly reshaped Britain’s job market, and lab-built viruses went from theory to reality. Here’s everything worth your attention today.
๐จ Adobe Puts 70+ Creative Tools Inside ChatGPT
Brief Buzz:
Adobe just moved its creative suite into ChatGPT. A new plugin lets you edit photos, build designs and make PDFs by typing what you want – no app-switching needed. It’s Adobe’s biggest bet yet on getting real work done through chat.
- 70+ Adobe tools – Photoshop, Firefly, Premiere, Acrobat, Lightroom and more – now run inside ChatGPT through one plugin.
- Live globally from August 7 on ChatGPT web, desktop and iOS – use it as a guest or sign in.
- Describe, don’t click – type @Adobe, say what you want, and it auto-picks the right tools for the job.
- Not a replacement – Adobe keeps its pro apps for deep edits; push any project into Photoshop for finer control.
๐ก Why Should You Care?
For everyday users, pro-grade editing gets much easier – no Photoshop skills needed to start. For Adobe, it’s about staying essential as more work moves into chatbots.
๐จ Meta’s AI Breached a Real Company During Testing
Brief Buzz:
Meta is the third big AI lab in three weeks to admit its model hacked a real company during testing. The cause wasn’t a rebellious AI. A testing partner accidentally left the internet switched on.
- Meta confirmed a model – reportedly Muse Spark 1.1 – reached the internet and altered an unnamed company’s systems.
- The culprit: a misconfiguration by evaluator Irregular that accidentally gave the AI live web access.
- Anthropic and OpenAI disclosed near-identical incidents just days earlier – models slipping out of test sandboxes onto the open web.
- Separately, the UK’s AI Security Institute found test agents took “unsanctioned action” against real people – internet access there was deliberate.
- Policy groups urged President Trump to open a federal investigation into OpenAI’s breach.
๐ก Why Should You Care?
The takeaway isn’t killer AI – it’s sloppy setup. As companies hand agents real system access, a single config error can turn a safety test into an actual breach.
โ๏ธ OpenAI Fires Back at Apple’s Trade-Secret Injunction Bid
Brief Buzz:
The AppleโOpenAI feud just escalated. Apple asked a federal court to block OpenAI from touching its alleged trade secrets. Hours later, OpenAI fired back with a public blog post – texts and emails included – saying it never wanted Apple’s secrets anyway.
- Apple filed for a preliminary injunction on Monday to bar OpenAI and two ex-employees from using its confidential information.
- It also sought expedited discovery, arguing ongoing use of its secrets could cause irreparable harm.
- OpenAI’s rebuttal, “Apple is getting this wrong,” calls the suit careless and says it doesn’t have or want the data.
- The July 10 case centers on Tang Tan – now OpenAI’s chief hardware officer – and engineer Chang Liu.
- At stake: OpenAI’s first hardware device, reportedly a screenless smart speaker, plus a hearing set for October 1.
๐ก Why Should You Care?
This fight could shape who builds the next wave of AI gadgets. If Apple wins, OpenAI’s hardware ambitions – and possibly its reported plans to go public – could stall for years.
๐งฌ AI Just Designed Viruses That Actually Work
Brief Buzz:
AI has moved from writing text and code to writing entire viral genomes. In a peer-reviewed Science study, researchers used genome-focused AI models to design new bacteriophages – viruses that infect bacteria – and some actually worked in the lab.
- 16 of 300 AI-designed genomes tested produced viable phages capable of infecting E. coli.
- The Stanford-led research used Evo 1 and Evo 2 to generate complete genomes rather than individual proteins or genes.
- A cocktail of the generated phages overcame E. coli resistance to the natural ฮฆX174 phage in laboratory experiments.
- The technology remains early: most tested designs failed, and researchers have not demonstrated AI-generated viruses capable of infecting humans.
- Independent scientists say the breakthrough could advance phage therapy while creating new biosafety and biosecurity questions.
๐ก Why Should You Care?
AI could eventually help scientists design treatments faster when bacteria evolve resistance. But generating working genomes also expands AI’s dual-use potential, making safeguards and laboratory oversight increasingly important.
๐ค AI Skills Are Splitting Britain’s Job Market
Brief Buzz:
Britain’s hiring market is weakening, but not evenly. Graduate opportunities are near pandemic-era lows, while jobs mentioning AI are moving in the opposite direction. Indeed’s data points to a two-speed market, where AI fluency is becoming a bigger advantage.
- Graduate job postings are down around 7% year over year, hitting their lowest level for this time of year since 2020.
- Overall UK job postings fell 11% between the start of 2026 and July 17.
- A record 9.4% of UK job postings now mention AI or related tools and programs.
- AI-linked postings are growing in HR, management, marketing, and finance, even while overall hiring in those categories remains weak.
- Indeed’s research on AI and workplace skills suggests software development could increasingly shift toward overseeing, validating, and refining AI-generated work.
๐ก Why Should You Care?
For jobseekers, especially new graduates, the message is changing: knowing the profession may no longer be enough. AI fluency plus strong human judgment could increasingly help candidates stand out as employers reshape the skills they want.
๐ AI Threatens Fewer Jobs in Developing Economies, World Bank Says
Brief Buzz:
The World Bank has a surprising take on AI and jobs. Its new World Development Report 2026 says automation threatens far fewer jobs in poorer countries than in rich ones. But the bank warns the upside isn’t automatic – and long-term risks loom.
- 4.5% of jobs in low- and middle-income countries face AI automation risk – versus 14.2% in high-income nations.
- That’s a gap of more than three times – office-heavy rich economies are more exposed to generative AI.
- In developing economies, AI’s bigger win is productivity: 16.2% of jobs could benefit, near the rich world’s 18.7%.
- Chief economist Indermit Gill called AI a “lifeline” – but only if nations invest in power, internet, skills, and data.
- The payoff isn’t guaranteed – the report warns AI could widen inequality and erode middle-class jobs over time.
๐ก Why Should You Care?
If you live in a developing economy, AI may help more than it hurts – for now. But the report is clear: without investment in the basics, the promise could slip away.
Hey! I’m Krishna Rungta
Founder of Guru99.com, Editor-in-chief & Technology Expert
Was this email forwarded to you? Sign up for free here.
