Sweden Restricts ChatGPT And Claude On Government Work Computers Over Data Security

by THEFUTURE.TEAM

October 26, 2026

Swedish government employees can no longer access browser-based versions of OpenAI’s ChatGPT or Anthropic’s Claude on work computers, as officials tighten controls over how sensitive information is handled.

Concerns Over Sensitive Data

The restriction follows a review of artificial intelligence tools used by staff at the Swedish Transport Administration earlier this year. The review identified what officials described as “problematic” uses of the platforms in government work.

Follow THE FUTURE on LinkedIn, Facebook, Instagram, X and Telegram

Officials were concerned that, without agreements specifying how data is protected, employees could enter sensitive information into services that might retain it or use it to train their models. The use of tools outside approved security arrangements—often referred to as shadow IT—could have “consequences for our information security,” said Mikael Wettercrantz, director of digitalisation at the Government Offices of Sweden.

Microsoft Copilot Recommended Under Security Agreement

The government is recommending Microsoft Copilot as an alternative. The service has passed a security screening process and is covered by an agreement between Microsoft and the Swedish government.

AI tools will remain available to government workers, but under tighter controls. A limited number of employees may continue using ChatGPT and Claude through a dedicated platform for AI assistants developed by Swedish authorities.

Restrictions Extend Beyond One Agency

According to Swedish public broadcaster Sveriges Radio, restrictions on ChatGPT and Claude have also been introduced at the Swedish Tax Agency and the National Board of Institutional Care. The Swedish Enforcement Authority is reviewing its approach.

Part Of A Wider Review Of AI Tools

ChatGPT and Claude join DeepSeek on the Swedish government’s list of AI tools that are not approved because of data-security concerns. The article also reports that the United States announced a ban on products and services from the Chinese AI company the previous week.

The decision highlights a central challenge for public-sector AI adoption: agencies must balance the potential efficiency gains of generative AI with clear safeguards for confidential information. For government organizations, approved tools and formal data-protection agreements are becoming essential conditions for use.

The Future Forbes Realty Global Properties

Blood Proteins May Signal Prostate Cancer Risk More Than 15 Years Before Diagnosis

by THEFUTURE.TEAM

October 27, 2026

Distinctive proteins in men’s blood may offer clues to prostate cancer risk more than 15 years before a diagnosis, according to research from the University of Oxford. The findings could help scientists better understand why some prostate cancers become aggressive while others grow slowly and never cause serious harm.

Researchers analyzed blood samples from 2,555 men collected between 1992 and 2000. Of those participants, 982 were later diagnosed with prostate cancer; the remaining 1,573 served as a comparison group. The team measured more than 6,000 proteins in the samples, which had been collected years—and, in some cases, more than 15 years—before diagnosis.

Follow THE FUTURE on LinkedIn, Facebook, Instagram, X and Telegram

Protein Patterns Were Linked To Different Risks

The analysis identified proteins associated with prostate cancer risk overall, as well as proteins linked specifically to more aggressive or fast-growing forms of the disease. Higher levels of certain immune-system proteins, including GZMK and ANXA2, were associated with a lower long-term risk.

“Finding proteins linked to known prostate cancer driver genes in blood collected years before diagnosis gives us new insights into the earliest stages of the disease,” said Karl Smith-Byrne, the study’s corresponding author at Oxford.

The results, published in European Urology, are significant because prostate cancer is not a single, uniform disease. Many tumors grow slowly and may never threaten a patient’s health, while others progress rapidly and can be life-threatening. A clearer way to distinguish between these outcomes could eventually help improve how clinicians assess risk.

Promising Findings, Not A Screening Test

The researchers stressed that the protein patterns are not ready to be used as a screening test. The associations need further validation before they can inform clinical decisions, and the study does not establish that the proteins cause cancer or that measuring them can predict an individual’s outcome.

“Ultimately, we hope this will help us understand the biology of aggressive prostate cancer and identify markers that could improve how we assess risk in the future,” said first author Zhe Huang, a postdoctoral epidemiologist at Oxford Population Health.

Research Could Point To New Prevention Strategies

Prostate cancer is the most commonly diagnosed cancer among men in the European Union and the second most common worldwide, after lung cancer. In the EU, approximately 330,000 cases are diagnosed and 75,000 men die from the disease each year.

By identifying biological signals that appear long before diagnosis, the research may help scientists investigate the pathways involved in aggressive disease. Smith-Byrne said that, alongside existing and future genetic evidence, the findings could help reveal pathways that might ultimately be targeted to prevent aggressive prostate cancer.

Breaking news