Anthropic launches Claude Sonnet 5.5 at the same price as Sonnet 5
On 28 September Anthropic released Claude Sonnet 5.5, the second model in the Claude 5.5 family. According to Anthropic, the model generates more than 30% faster than Sonnet 5 and is especially strong at well-defined everyday tasks, fixing bugs, and creating documents, presentations and spreadsheets. The API price remains $2 per million input tokens and $10 per million output tokens; the model is also available via AWS, Google Cloud and Microsoft Azure, with zero data retention.
What does this mean? If you use Sonnet 5 via the API or a tool built on it, you get faster responses without a higher token price; test the new model on your own documents before you switch.
Advocate General at the Supreme Court: ChatGPT may be used for damage calculation, but must be verifiable
In a criminal case the victim had estimated the value of stolen jewellery (€ 49,874) using ChatGPT, and the court adopted that estimate. Advocate General Paridaens writes that no rule of law precludes the use of an AI tool such as ChatGPT, but that the calculation must be sufficiently transparent and verifiable for the other parties and the judge. Because the court gave insufficient reasons for the award, the Advocate General advises quashing the judgment on that point; the Supreme Court has yet to decide.
What does this mean? If you support a claim or damages item with help from AI, make sure every assumption and figure can be traced to its own verifiable source.
Google replaces Gemini Gems with skills from 17 November
From 17 November 2026, Google will automatically convert Gems in the Gemini app, custom assistants with fixed instructions, into skills. You call skills with a "/" in the input field and you can use several in one conversation. Skills are currently only available to Google AI Pro and AI Ultra subscribers; it is still unclear what happens for users with a free account.
What does this mean? If you have work processes set up around Gems, check after 17 November that your instructions have been transferred correctly and whether your subscription gives you access to skills.
Philips research: healthcare staff use their own AI tools without training
Nearly two-thirds of healthcare workers turn to their own AI tools such as ChatGPT, according to research commissioned by Philips, reported by De Telegraaf. 83% say training is lacking, limited in availability or not structurally offered, and more than half lack clear processes for checking AI use. Just over a quarter have AI suggest diagnoses based on symptoms.
What does this mean? If you work in healthcare, don't enter patient data into a personal AI tool, and ask your organisation which tools are permitted.
