Synaptics threatens hobbyist over Linux driver he built with Claude

What's going on

Hardware maker Synaptics is going after a hobbyist developer who wrote a free Linux driver for DisplayLink devices. DisplayLink chips are inside many USB docking stations. They let you connect extra monitors over USB and show up in products like Dell's D6000. Synaptics acquired DisplayLink in July 2020.

The developer, Mike Lothian, calls his driver Vino. According to Linux news site Phoronix, he built it through clean-room reverse engineering, partly with AI. Tom's Hardware reports that he used a subscription to Anthropic's Claude. Synaptics sent him cease-and-desist letters and threatened to sue. Lothian has since taken the GitHub repository down, but he isn't backing off. He has posted the entire email exchange to the Linux kernel mailing list.

What Vino could do

Until now, using DisplayLink docks on Linux has meant two pieces of software: a kernel module called EVDI, which isn't part of the mainline Linux kernel, and a background service called DisplayLinkManager, which is closed source. According to MLQ.ai, Vino was meant to replace both. It's written in Rust as a DRM/KMS driver, so it uses the standard interface Linux relies on to drive displays. The goal was for the docks to behave like any other Linux monitor.

By late August, MLQ.ai says, Vino had displays working on three dock families:

"Clean room" means the functionality is rebuilt purely by observing and analyzing how the device behaves, without looking at anyone else's source code. Legally, that matters: it undercuts any claim that code was copied.

How the dispute escalated

According to Tom's Hardware, it all started innocently in July 2026. Lothian asked Łukasz Spintzyk, a software engineer at Synaptics, for help with a few bugs. A few weeks later, the reply came from someone else: Michal Lukaszek, who handles strategic engineering partnerships at Synaptics. The company was concerned that Lothian was implementing the DisplayLink protocol himself and infringing its patents. A letter from the legal department followed, demanding among other things that he delete the GitHub repository. The correspondence on the kernel mailing list mentions cease-and-desist letters dated August 17 and October 5, 2026.

Synaptics explicitly does not accuse Lothian of copyright infringement. This is purely about patents, specifically two UK ones:

• GB2484736 ("Image generation"): describes how display updates are sent over a general-purpose data network such as USB or Ethernet using a particular architecture and compression method.
• GB2485613 ("Display control device"): describes hardware that works with compressed data and only decompresses it when needed.

Lothian asked for a detailed breakdown of how his driver supposedly infringes these patents. When he got no answer, he followed up three weeks later. In response, Andrew Bateman, Synaptics' deputy general counsel, said the company would send him an invoice "to cover the costs and effort required to document your infringing activities." According to Tom's Hardware, he also wrote:

"By open-sourcing our DisplayLink technology as a 'hobbyist,' for whatever personal reasons, you are single-handedly destroying DisplayLink's competitive advantage in this competitive market."

And further:

"If you wish to inform the media or otherwise publish our response, I encourage you to do so. We welcome any free publicity that shows the world – especially our competitors and shareholders – that Synaptics aggressively defends its DisplayLink intellectual property."

According to MLQ.ai, Synaptics has not commented on the case publicly.

Lothian's counterarguments

On October 14, Lothian replied to Clive So, a partner at London patent firm J A Kemp, which represents Synaptics. "I do not accept that any of my actions infringe either of the patents," he wrote. He makes three arguments:

1. The patents are invalid. In both cases, he says, the claims as originally filed had nothing to do with networking or transmission. The patents were only granted after transmission over a "general-purpose data network" was added, and in the case of GB2485613 also a "transport protocol" with a decompression step. But that addition wasn't new. As prior art (earlier publications that anticipate an invention), he cites the 1998 RFB/VNC specification for remote access to screens.
2. The patented processes run inside the dock itself. Users bought those docks perfectly legally.
3. Open Invention Network (OIN): Synaptics is a member of this patent pool, whose members commit not to use their patents against Linux. Lothian argues that this also covers patents held by subsidiaries like DisplayLink. He has been an OIN licensee himself since August 17. Synaptics apparently disagrees and says DisplayLink isn't part of the OIN. Lothian asked the law firm to identify the specific clause or "Limitation Election" that excludes the DisplayLink patents. He also said any case would belong before the Court of Session in Scotland.

According to Tom's Hardware, Lothian remains open to a reasonable settlement.

AI as a new factor in reverse engineering

The case is drawing extra attention because AI made it possible in the first place. Writing a kernel driver for an undocumented protocol used to be a job for specialists with plenty of time. With tools like Claude, individuals can now do it in their spare time. Tom's Hardware cites other examples: one user reportedly used Claude to build a free clone of Adobe Creative Suite in Rust and is said to be working on a free office suite and Autodesk-style tools as well. Many reverse-engineered games are also being ported to other platforms. According to Tom's Hardware, this prompted Anthropic to introduce certain restrictions in Opus 5.5. The report doesn't go into detail.

Our take

This case matters for two reasons. For Linux users with DisplayLink docks, Vino would have been a real step forward: a free driver that could be integrated into the kernel instead of a closed add-on. More broadly, it shows that companies are increasingly turning to patents to fight AI-assisted clones when copyright no longer applies because no code was copied.

Whether Synaptics' claims hold up is completely open. No court has looked at them. Lothian's arguments on prior art and the OIN license at least don't sound far-fetched, but they haven't been tested either. Many observers are critical of how Synaptics has handled this: threatening to charge an individual just for explaining the allegations against him is intimidation, not argument. That's exactly the debate in the reader comments at Tom's Hardware, including references to US laws against intimidation lawsuits (anti-SLAPP). Patent law doesn't come with a hobbyist discount. A long legal battle can ruin an individual even if he wins in the end. That's probably the real reason the repository is offline for now.