English

OpenAI has recently introduced GPT-6 Astra, describing it as an important step toward AGI, or Artificial General Intelligence. AGI refers to an AI system that can perform many different intellectual tasks, learn from situations, adapt to unfamiliar problems, and work flexibly across different fields. However, there is no universally accepted definition or official test for AGI, so claims that we have entered the “AGI era” remain open to debate. What makes GPT-6 Astra different from earlier chatbots is its ability to work toward a complete objective rather than simply answer one question at a time. Users can give Astra a general goal without explaining every step. The system can then understand the objective, create a plan, divide the work into smaller tasks, and continue until it produces a final result. Astra is also described as being able to interact with a laptop, use software, browse websites, write and test code, and control digital tools. For example, a user might ask it to create a website, prepare a presentation, or build a software project. Astra can potentially plan the work, complete the different stages, check for errors, make revisions, and deliver the finished product. This is where the idea of an AI agent becomes important. An AI agent is a system that can receive an objective, make decisions, use tools, take actions, review its progress, and continue working with limited step-by-step guidance. In simple terms, a chatbot usually tells you how to do something, while an agent attempts to do it for you. For example, instead of asking Astra how to create a stress-management workbook, a therapist could simply provide the objective: “Create a professional workbook for university students.” Astra could organize the content, write the exercises, design the document, review the language, correct problems, and prepare a final version. The therapist would still need to check the material before using it. GPT-6 Astra can be compared with Claude Fable 5, although their strengths may be different. Astra appears more focused on autonomous task completion, computer use, coding, and interaction with digital tools. Claude Fable 5 may be viewed more as a strong writing, reasoning, and analytical partner. In practice, Astra may be better suited to completing a project from beginning to end, while Fable 5 may be particularly useful for developing ideas, analyzing information, and producing carefully written content. Feature GPT-6 Astra Claude Fable 5 Main strength Autonomous task completion Writing and reasoning Typical use Coding, computer use, and digital projects Analysis, idea development, and content creation Working style Completes projects from beginning to end Supports users through careful discussion and drafting Best suited for Complex, multi-step tasks Research, writing, and critical analysis For therapists, these developments could have practical benefits. AI agents might help organize research, prepare psychoeducational materials, create therapy exercises, summarize non-identifying information, support administration, and develop resources for clinical training or rehabilitation. By reducing routine workload, these tools may give therapists more time for direct patient care. However, an AI’s ability to create software or solve complex problems does not mean that it understands patients in a human way. It may produce an impressive handout while misunderstanding a patient’s culture, emotional state, personal history, or level of risk. Therapy depends on empathy, trust, communication, and careful interpretation, which remain central to safe clinical practice. Greater independence also creates new responsibilities. Therapists must know what the AI agent can access, which websites or files it can use, and when human approval is required. An AI should not independently diagnose a patient, assess suicide risk, change treatment, or communicate sensitive clinical information. Its work must always be reviewed before it influences patient care. So, are we entering the AGI era? GPT-6 Astra may represent a significant movement toward AI systems that can plan, act, revise, and complete complex objectives independently. Whether this should officially be called AGI remains uncertain. For therapists, the safest approach is to use such systems as powerful assistants while keeping privacy, professional judgment, empathy, and patient safety at the centre of clinical practice.

Children are growing up in a world where chatbots and other AI tools are part of everyday life. They may use AI to ask questions, complete homework, create stories, solve puzzles, or imagine the future. For therapists, the important question is not simply how often children use AI. We also need to understand how these tools may influence children’s learning, emotions, relationships, and sense of identity. Children develop through active engagement with the world around them. Exploring, playing, communicating, moving, making mistakes, solving problems, and building relationships are essential experiences for developing cognitive, social, emotional, motor, and problem-solving skills. These processes require time, effort, curiosity, and direct interaction with other people and the environment. Generative AI can provide immediate answers and produce stories, images, homework, and solutions. Although this may seem helpful, it can become problematic when AI replaces opportunities for independent thinking, creativity, exploration, and learning through experience. Children may begin to rely on technology rather than developing confidence in their own abilities and ideas. There are also concerns about children’s ability to evaluate AI-generated information. Chatbots can provide inaccurate or misleading answers while communicating in a confident and friendly manner. Young children may not yet have the critical-thinking skills needed to recognise these limitations. Some may also view chatbots as friends, teachers, or trusted companions, even though AI does not genuinely understand their emotions or experiences. For these reasons, direct access to generative AI should not be considered appropriate for young children. Early learning should primarily occur through imaginative play, real-world exploration, human relationships, outdoor activities, and hands-on experiences. These experiences cannot be replaced by instant digital interaction. For teenagers, generative AI may be introduced more cautiously as a supported tool. Its use should involve parental or educational guidance, clear boundaries, privacy protection, and ongoing attention to critical thinking. Teenagers should be encouraged to use AI to support learning, not to replace effort, creativity, communication, or problem-solving. Just as we have age-based recommendations for screen use, we should also ask: At what age is generative AI appropriate for children, and under what conditions? Clear, evidence-based recommendations regarding appropriate ages, limits, and safe use are still needed. Until stronger guidance is available, adults should take a cautious approach. Children’s development, wellbeing, creativity, and relationships must come before convenience and technological progress. AI should never replace the human experiences that help children learn, grow, and understand themselves and the world around them. Coming Soon! We’re preparing something special for parents! Webinar: Generative AI and Children & Teenagers A practical discussion exploring when AI may be appropriate, how it can affect learning and development, and how parents can introduce it safely and responsibly. More details coming soon, stay tuned!

Jais is an open-weight artificial intelligence (AI) model created especially for Arabic. It can understand and produce Arabic, English, different Arabic dialects, Arabizi (Arabic written with Latin letters and numbers), and some cultural forms such as poetry and religious language. This is important because many AI tools work better in English than in Arabic. Jais may be useful in healthcare, education, research, and therapy. For example, therapists could use it to translate information, simplify difficult explanations, prepare educational materials, or create examples that are more familiar to Arabic-speaking clients. It may also help researchers work with Arabic-language information. However, Jais should be used as a support tool, not as a replacement for professional knowledge. A therapist should never depend on the model to make a diagnosis, assess risk, or decide on treatment. AI can produce incorrect or incomplete answers, even when the response sounds confident and professional. Arabic-speaking communities are very diverse. People may use different dialects, cultural expressions, and ways of describing emotions or illness. Although Jais is designed to work well with Arabic, it may still misunderstand a phrase or make assumptions about a person’s culture, religion, or background. Human review is always necessary. The fact that Jais is “open-weight” means that researchers and developers can study and adapt the model more easily than some closed commercial systems. This may support local research and the development of Arabic healthcare tools. However, open access does not guarantee that the model is always accurate, unbiased, or transparent. Therapists should begin with simple and low-risk uses. For example, they may ask Jais to rewrite a verified explanation in easier Arabic or English. They should then check the result carefully before sharing it with a client. Personal or confidential patient information should not be entered unless privacy and security have been officially confirmed. Ethical responsibility remains with the professional using the tool. Therapists and researchers should check AI-generated information, protect patient privacy, and be transparent about how AI was used. They should also consider possible bias, differences between dialects, and the risk that some groups may receive less accurate responses. Jais is a promising development for Arabic-centered AI, but its value will depend on responsible use and continued evaluation. With careful human supervision, it may improve communication and access to information. It should help professionals support Arabic-speaking communities—not replace clinical judgment, human relationships, or accountability.

For many Deaf people, communicating with hearing people by phone can be difficult. Typing may take time, and written English may not be their first language. Google DeepMind is developing SL2T, an artificial intelligence model that can recognise American Sign Language through a phone camera and translate it into written English. This technology could support everyday communication in many situations, including appointments, education, work, travel, shopping, and contact with public services. Patients may also be able to use it independently in their daily lives to communicate with family members, friends, colleagues, customer service workers, and other people who do not know sign language. This could reduce reliance on others and give Deaf users more control over how they communicate. In healthcare and therapy, SL2T may help patients describe simple needs, understand information, ask questions, and communicate more independently with therapists and other professionals who do not know sign language. It could also help Deaf people communicate with the wider community and society, including neighbours, shop workers, teachers, employers, transport staff, and public service providers. This may support greater participation in daily life and help reduce communication barriers. However, AI translation should be viewed as a communication aid, not a replacement for a qualified interpreter. Sign languages have their own grammar, and meaning can depend on hand movements, facial expressions, body position, speed, and context. If the system misses one of these elements, the translation may be inaccurate. Therapists should therefore use this technology cautiously. It may be suitable for short messages, basic questions, appointment information, or informal communication. It should not be used alone for diagnosis, informed consent, mental health risk assessment, safeguarding, or important treatment decisions. In complex or sensitive situations, a professional interpreter remains essential. The system may also have difficulty with regional signs, rapid signing, fingerspelling, poor lighting, or unusual camera angles. A translation may appear clear but still be wrong. Professionals should check important information with the patient and give them opportunities to correct the result. The involvement of Deaf people in developing and testing the technology is an important strength. Their experience can help developers understand real communication needs and avoid treating sign language as simply a technical problem. Further research is still needed across different sign languages, cultures, regions, and healthcare settings. Privacy is another concern. Signing may communicate personal, medical, or emotional information. Therapists should explain how the tool works, whether recordings are stored, and who can access the data. Patients should always be free to choose whether they want to use it. AI can improve accessibility, but professional responsibility remains essential. Therapists must verify the information, recognise possible bias, and avoid assuming that the system is always correct. Used carefully, SL2T may give Deaf users greater independence and more communication choices while supporting, not replacing, human understanding.

Since August 2, 2026, many professionals have been discussing Article 50 of the European Union’s AI Act. This article introduces new transparency rules for certain uses of all types of artificial intelligence, not only generative AI. However, these rules do not mean that therapists must inform patients every time they use an AI tool. The most important point is to distinguish between two different situations. The first situation is when the patient communicates directly with an AI system. For example, a clinic may use a chatbot to answer questions, help families find information, or arrange appointments. In this case, the patient should normally be told that they are interacting with artificial intelligence rather than with a human professional. This information should be clear and easy to understand. The second situation is when the therapist uses AI privately as a work tool. For example, a therapist may use AI to improve the wording of a report, organize information, create ideas for therapy materials, or structure a document. When the patient does not communicate directly with the AI, Article 50 does not create a general obligation to tell the patient about every use of the tool. However, using AI in the background does not remove the therapist’s responsibility. AI can make mistakes, misunderstand information, or produce answers that sound convincing but are incorrect. Any text or suggestion created by AI must therefore be carefully read, checked, and adapted by the professional before it is used. It is also important to distinguish between writing support and clinical decision-making. Asking AI to improve the wording of a sentence is different from asking it to interpret test results, suggest a diagnosis, or recommend treatment. In the second case, AI becomes much more involved in clinical reasoning. This requires greater caution and should never replace the therapist’s own professional judgment. Clinical decisions must be based on the patient’s complete situation, including their history, test results, communication, environment, culture, and individual needs. AI does not always have access to this context. It may help generate ideas, but the therapist remains responsible for deciding what is clinically appropriate. The AI Act also does not replace the General Data Protection Regulation, confidentiality rules, or professional ethics. Before entering patient information into an AI tool, professionals should ask whether the data can legally and safely be shared. Names, health information, test scores, clinical reports, and identifiable details should not be entered into public AI tools without appropriate protection. For therapists, “AI literacy” does not simply mean learning how to write good prompts. It means understanding when AI may be useful, when it should not be used, how to check its answers, how to protect patient data, and how to recognize possible errors or bias. A simple rule can help: If the patient communicates directly with AI, transparency is needed. If the therapist uses AI in the background, careful supervision, data protection, and professional validation are essential. AI can support clinical work, but it cannot replace the therapist’s expertise, responsibility, or relationship with the patient. The goal is not to avoid every AI tool, but to use these technologies carefully, transparently, and in a way that protects patients and maintains the quality of professional care. References Artificial Intelligence Act. (n.d.). Article 50: Obligations de transparence pour les fournisseurs et les déployeurs de certains systèmes d’intelligence artificielle. Retrieved August 25, 2026, from https://artificialintelligenceact.eu/fr/article/50/

Higgsfield.ai is now available inside ChatGPT, allowing users to create images and videos without moving between different applications. Through a conversation, users can develop visual materials, short videos, voiceovers, and captions. The system can also connect with creative models such as Seedance, Seedream, GPT Image, and Kling, as well as tools for narration, subtitles, and faceless video creation. One important advantage is that therapists do not need to learn how to use Higgsfield.ai or understand the differences between all the available models. Instead of choosing between different video, image, or narration tools, users can simply describe what they want and provide a clear prompt. ChatGPT can then help select the most appropriate model or option for the requested result. This makes the creative process easier and more accessible, especially for therapists who do not have a technical or video-editing background. For therapists, this may be useful because many patients understand information more easily when it is presented visually. For example, a therapist working with a child could create a short story about identifying emotions, calming down, or solving a social problem. With an adult patient, the therapist might create a simple visual explanation of anxiety, stress, sleep difficulties, or relaxation techniques. These tools may also support patients who have communication, learning, attention, or memory difficulties. A visual sequence could show the steps of a daily routine. A narrated video could explain a coping strategy. Subtitles may help patients who have hearing difficulties or who understand written information better than spoken instructions. The material can be adapted to the patient’s age, language level, culture, and therapeutic goals. However, the quality of the technology does not automatically make the material clinically useful. A video may look professional but still contain confusing, inaccurate, or culturally inappropriate content. Therapists should review every image, video, voiceover, and caption before sharing it with a patient. The material should support therapy, not replace clinical explanation, professional judgment, or the therapeutic relationship. Higgsfield.ai may also help therapists prepare psychoeducational materials more quickly. A therapist could ask ChatGPT to suggest a simple script, create a visual metaphor, or design a short educational video. This may be especially helpful when preparing materials for children, families, groups, or patients with different learning needs. Still, therapists should use these tools as assistants rather than as independent decision-makers. Privacy is one of the most important concerns. Therapists should not enter patient names, clinical notes, identifiable photographs, voice recordings, or private personal information into these systems unless the service has been approved by their organization and meets relevant privacy requirements. When possible, therapists should use general examples and fictional characters instead of real patient information. It is also important to tell patients when a resource has been created with AI. This promotes honesty and helps patients understand that the material is a digital support, not a human-created clinical opinion. Therapists remain responsible for checking the accuracy, safety, and appropriateness of anything they use. AI cannot take responsibility for diagnosis, treatment decisions, informed consent, or possible harm. Overall, Higgsfield.ai inside ChatGPT may make it easier for therapists to create accessible and engaging materials. Its greatest value is not simply producing attractive videos or images, but helping therapists explain ideas in ways that patients can see, hear, and understand. Used carefully, transparently, and with strong attention to privacy, it can become a useful addition to clinical practice rather than a replacement for professional expertise.

Recent reports suggest that OpenAI’s new AI system, Astra, may have solved several difficult mathematical problems. These solutions were checked by Lean, a computer program that verifies each step of a mathematical proof. This development raises an important question for therapists: if AI can solve problems that humans cannot, could it also make better decisions about diagnosis and treatment? AI may become a useful support tool in health care. It can quickly examine large amounts of information, such as symptoms, research studies, treatment results, and patient records. In therapy, AI could help identify possible patterns, suggest questions for further assessment, or remind clinicians to consider conditions that may otherwise be missed. It may also help therapists plan more personalised interventions. However, diagnosis in therapy is not only a technical or mathematical task. Clients are complex individuals. Their emotions and behaviours may be influenced by trauma, family relationships, culture, language, stress, physical health, disability, and social circumstances. Two people with similar symptoms may need completely different support. AI can identify patterns, but it may not fully understand the personal meaning behind those patterns. AI may suggest a possible diagnosis or the treatment that appears most successful in research. However, this suggestion should not replace the therapist’s professional judgement. AI can make mistakes when information is incomplete, inaccurate, or biased. It may also misunderstand cultural differences or interpret normal reactions to difficult life events as signs of mental illness. The therapist’s role will therefore change, but it will not disappear. Therapists may use AI to organise information and support clinical decision-making. They will still need to check whether the information is correct, explain it to clients, and decide whether it is suitable for the individual person. Therapists will also remain responsible for listening carefully, showing empathy, building trust, and adapting treatment when the client’s needs change. The therapeutic relationship is one of the most important parts of care. Clients often improve because they feel safe, respected, and understood. AI may analyse words, behaviours, or clinical data, but it cannot replace genuine human presence and connection. A computer may suggest a treatment, but the therapist helps the client understand it, accept it, and use it in daily life. There are also important ethical concerns. AI systems may be trained using data that do not properly represent all cultures, languages, ages, or disabilities. Their recommendations may be difficult to explain. Therapists must protect client privacy and avoid sharing personal information without proper safeguards. They must also remain accountable for decisions made with the help of AI. The future of therapy will probably involve cooperation between humans and AI. AI may become a powerful assistant that helps therapists manage information and keep up with research. However, it should support, not replace, professional judgement. Therapists will remain responsible for understanding the client as a whole person and making decisions together with them. AI may solve complex mathematical problems, but therapy involves more than finding a correct answer. It requires compassion, communication, ethical judgement, and an understanding of human experience. The future role of therapists will be to use AI carefully while protecting the human relationship at the centre of care.

Artificial intelligence (AI) is increasingly used in health research. It can analyse medical records, images, speech, and other health information. It can also support literature reviews, writing, data analysis, diagnosis, and treatment planning. Although AI can improve research and health care, it also creates important ethical questions about privacy, fairness, safety, and human responsibility. For example, an AI system may be developed to identify depression from a person’s speech. Even if the system appears accurate, researchers must ask whether participants understood how their recordings would be used. They must also consider whether the system works equally well for different languages, cultures, ages, and disabilities. These questions show why AI research requires careful ethical review. The World Health Organization (WHO, 2026) explains that AI research should be reviewed throughout the study, not only at the beginning. AI systems may change when new data are added, and their results may be difficult to explain. Researchers should continue checking the system for errors, unexpected risks, privacy problems, and differences between research results and real clinical practice. Basic ethical principles remain important. Researchers should respect people, avoid harm, provide possible benefits, and treat groups fairly. They should clearly explain how health data will be collected, stored, shared, and used. They must also protect participants from discrimination, stigma, incorrect information, and unsafe clinical decisions. Fairness is a major concern. An AI system may work well for one group but poorly for another. For example, a system trained mainly with data from English-speaking adults may not work well for children, older adults, people with disabilities, or people from other cultural backgrounds. Researchers should use diverse data and report clearly where the system performs well and where it has limitations. These issues are especially important in low- and middle-income countries. Data may be collected from local communities but used by organisations in other countries. The communities that provide the data may receive few benefits or have little control over its use. Ethical research should include local researchers, community involvement, fair benefit sharing, training, and long-term capacity-building. Researchers should also describe clearly how AI was used. They should report the tool, its purpose, the data involved, and the methods used to check errors. Poor-quality or incomplete data can lead to unfair or unsafe results. Therefore, researchers should examine not only overall accuracy but also safety, usefulness, transparency, and performance across different groups. AI may help ethics committees identify privacy risks, missing information, or possible bias. However, it should not replace human judgement. Ethical review requires understanding people’s rights, values, cultures, and experiences. Therapists and other health professionals should ask how an AI tool was developed, whether it was tested with their patients, and who is responsible when it makes a mistake. AI may support better research, earlier diagnosis, and improved health services, but its benefits are not automatic. Responsible use requires privacy protection, fairness, transparency, high-quality data, and continued human oversight. The important question is not only whether AI works, but whether it works safely and fairly while supporting, not replacing, professional judgement and compassionate care. References World Health Organization. (2026). Artificial intelligence-related health research: Ethics review and oversight. World Health Organization.

Generative artificial intelligence is becoming increasingly relevant to healthcare, offering therapists new ways to prepare clinical resources, organize ideas, and support individualized intervention. Tools such as ChatGPT can assist with therapeutic activities, caregiver education, professional communication, and session planning. However, AI should support, not replace, clinical expertise, ethical judgment, and the therapeutic relationship. We are pleased to introduce a practical AI training designed for therapists and healthcare professionals. The programme will explore how to use generative AI safely and effectively in therapy, with a focus on real clinical applications. Participants will learn how to integrate AI into their professional workflow while maintaining evidence-based practice and individualized care. A key component of the training is prompt engineering for therapists. Participants will learn how to create clear, specific prompts that produce more relevant and personalized results. They will also explore how to create therapeutic materials for children, teenagers, and adults, including activities, visual supports, social narratives, emotional-regulation resources, and caregiver-friendly explanations adapted to different clinical needs. Ethics and confidentiality are central to responsible AI use. The training will address patient-data protection, professional boundaries, informed decision-making, and HIPAA compliance where applicable. Therapists will learn how to reduce privacy risks, recognize inaccurate or biased information, and critically review all AI-generated content before using it in clinical practice. The programme is suitable for speech therapists, occupational therapists, physical therapists, psychologists, and other healthcare professionals. Through practical examples and clinical case applications, participants will develop the knowledge and confidence to decide when AI is useful, when content requires modification, and when professional judgment must take priority. Participants will receive 11 CPD points, with the programme certified or recognized by the UAE Department of Health. The training is designed to help clinicians use AI safely, wisely, and effectively while keeping confidentiality, accountability, and patient wellbeing at the centre of practice. Generative AI may become a valuable part of modern clinical workflows, but its success will depend on responsible human oversight. Therapists must question, verify, and adapt every AI-generated response through their own professional knowledge. For clinicians ready to take the next step into AI-powered therapy, this training offers a practical and ethical foundation for meaningful innovation.

Microsoft has made a new 3D-generation system available as open source. This means that researchers, developers, and studios can access the model, its code, and its training information. They can run it locally and adapt it for different projects. The system can create three-dimensional objects from images and export them for use in programs such as Blender and Unreal Engine. The system is also very fast. Microsoft reports that it can create a 5123 3D object (a highly detailed 3D model made using a grid of 512 × 512 × 512 small points, called voxels)  in about three seconds when using a powerful NVIDIA H100 computer. Larger and more detailed objects may take between 17 and 60 seconds. However, this technology requires expensive equipment, Linux, and an NVIDIA graphics card with at least 24 GB of memory. For therapists, 3D generation may provide new ways to create visual and interactive materials. An occupational therapist, for example, could create a model of a chair, kitchen tool, wheelchair, or bathroom. This model could help patients understand safety issues, practise daily activities, or discuss changes to their home environment. Three-dimensional models may be especially helpful for patients who learn better through pictures and objects than through spoken explanations. Therapists could use them to demonstrate how to hold an object, move through a room, or use an assistive device. This may support children and people with communication, cognitive, or learning difficulties. The technology could also support education and research. Students and professionals might create virtual homes, workplaces, or community spaces to study accessibility, balance, mobility, attention, and problem-solving. However, a realistic-looking model is not always accurate. AI may create objects with incorrect sizes, shapes, or details, so therapists must check every model before using it. Open-source access may allow researchers to study the system more carefully and explain how it was used. Nevertheless, open source does not mean that the technology is automatically safe or clinically tested. Results depend on the image, instructions, and editing process. Privacy is also important when using patient photographs, home images, or body scans. Therapists remain responsible for all professional decisions. AI should support clinical judgment, not replace it. Professionals should record how each model was created, explain AI use when appropriate, protect patient information, and make sure the final material is understandable and suitable for the patient. Microsoft’s 3D-generation system may become useful in therapy, education, and research. Its value lies in helping professionals create personalised and interactive materials more easily. Safe use will require careful checking, privacy protection, clear documentation, and further research into its benefits and limitations.