Author: azeeadmin

The Bundeskartellamt, Germany’s very active competition authority, isn’t letting the grass grow under new powers it gained this year to tackle big tech: The Federal Cartel Office (FCO) has just announced a third proceeding against Google.

The FCO’s latest competition probe looks very interesting as it’s targeting Google News Showcase — Google’s relatively recently launched product which curates a selection of third party publishers’ content to appear in story panels on Google News (and other Google properties), content for which the tech giant pays a licensing fee.

Google started cutting content licensing deals with publishers around the world for News Showcase last year, announcing a total pot of $1BN to fund the arrangements — with Germany one of the first markets where it inked deals.

However its motivation to pay publishers to licence their journalism is hardly pure.

It follows years of bitter accusations from media companies that Google is freeloading off their content. To which the tech giant routinely responded with stonewalling statements — saying it would never pay for content because that’s not how online aggregation works. It also tried to fob off the industry with a digital innovation fund (aka Google News Initiative) which distributes small grants and offers free workshops and product advice, seeking to frame publishers’ decimated business models as a failure of innovation, leaving Google’s adtech machine scot free to steamroller on.

Google’s stonewalling-plus-chicken-feeding approach worked to stave off regulatory action for a long time but eventually enough political pressure built up around the issue of media business models vs the online advertising duopoly that legislators started to make moves to try to address the power imbalance between traditional publishers and intermediating tech giants.

Most infamously in Australia, where lawmakers passed a news media bargaining code earlier this year.

Prior to its passage, both Facebook and Google, the twin targets for that law, warned the move could result in dire consequences — such as a total shut down of their products, reduced quality or even fees to use their services.

Nothing like that happened but lawmakers did agree to a last minute amendment — adding a two-month mediation period to the legislation which allows digital platforms and publishers to strike deals on their own before having to enter into forced arbitration.

Critics say that allows for the two tech giants to continue to set their own terms when dealmaking with publishers, leveraging market muscle to strike deals that may disproportionately benefit Australia’s largest media firms — and doing so without any external oversight and with no guarantees that the resulting content arrangements foster media diversity and plurality or even support quality journalism.

In the EU, lawmakers acted earlier — taking the controversial route of extending copyright to cover snippets of news content back in 2019.

Following on, France was among the first EU countries to transpose the provision into national law — and its competition watchdog quickly ordered Google to pay for news reuse back in 2020 after Google tried to wiggle out of the legislation by stopping displaying snippets in the market.

It responded to the competition authority’s order with more obfuscation, though, agreeing earlier this year to pay French publishers for linking to their content but also for their participation in News Showcase — bundling required-by-law payments (for news reuse) with content licensing deals of its own devising. And thereby making it difficult to understand the balance of mandatory payments vs commercial arrangements.

The problem with News Showcase is that these licensing arrangements are being done behind closed doors, in many cases ahead of relevant legislation and thus purely on Google’s terms — which means the initiative risks exacerbating concerns about the power imbalance between it and traditional publishers caught in a revenue bind as their business models have been massively disrupted by the switch to digital.

If Google suddenly offers some money for content, plenty of publishers might well jump — regardless of the terms. And perhaps especially because any publishers that hold out against licensing content to Google at the price it likes risk being disadvantaged by reduced visibility for their content, given Google’s dominance of the search market and content discoverability (via its ability to direct traffic to specific media properties, such as based on how prominently News Showcase content is displayed, for example).

The competition implications look clear.

But it’s still impressive that the Bundeskartellamt is spinning up an investigation into News Showcase so quickly.

The FCO said it’s acting on a complaint from Corint Media — looking at whether the announced integration of the Google News Showcase service into Google’s general search function is “likely to constitute self-preferencing or an impediment to the services offered by competing third parties”.

It also said it’s looking at whether contractual conditions include unreasonable terms (“to the detriment of the participating publishers”); and, in particular, “make it disproportionately difficult for them to enforce the ancillary copyright for press publishers introduced by the German Bundestag and Bundesrat in May 2021” — a reference to the transposed neighbouring right for news in the EU copyright reform.

So it will be examining the core issue of whether Google is trying to use News Showcase to undermine the new EU rights publishers gained under the copyright reform.

The FCO also said it wants to look at “how the conditions for access to Google’s News Showcase service are defined”.

Google launched the News Showcase in Germany on October 1 2020, with an initial 20 media companies participating — covering 50 publications. Although more have been added since.

Per the FCO, the News Showcase ‘story panels’ were initially integrated in the Google News app but can now also be found in Google News on the desktop. It also notes that Google has said the panels will soon also appear in the general Google search results — a move that will further dial up the competition dynamics around the product, given Google’s massive dominance of the search market in Europe.

Commenting on its proceeding in a statement, Andreas Mundt, president of the Bundeskartellamt, said: “Cooperating with Google can be an attractive option for publishers and other news providers and offer consumers new or improved information services. However, it must be ensured that this will not result in discrimination between individual publishers. In addition, Google’s strong position in providing access to end customers must not lead to a situation where competing services offered by publishers or other news providers are squeezed out of the market. There must be an adequate balance between the rights and obligations of the content providers participating in Google’s programme.”

Google was contacted for comment on the FCO’s action — and it sent us this statement, attributed to spokesperson, Kay Oberbeck:

“Showcase is one of many ways Google supports journalism, building on products and funds that all publishers can benefit from. Showcase is an international licensing program for news — the selection of partners is based on objective and non-discriminatory criteria, and partner content is not given preference in the ranking of our results. We will cooperate fully with the German Competition Authority and look forward to answering their questions.”

The FCO’s scrutiny of Google News Showcase, follows hard on the heels of two other Google proceedings it opened last month, one to determine whether or not the tech giant meets the threshold of Germany’s new competition powers for tackling big tech — and another examining its data processing practices. Both remain ongoing.

The competition authority has also recently opened a proceeding into Amazon’s market dominance — and is also looking to extend another recent investigation of Facebook’s Oculus business, also by determining whether the social media giant’s business meets the threshold required under the new law.

The amendment to the German Competition Act came into force in January — giving the FCO greater powers to proactively impose conditions on large digital companies who are considered to be of “paramount significance for competition across markets” in order to pre-emptively control the risk of market abuse.

That it’s taking on so many proceedings in parallel against big tech shows it’s keen not to waste any time — putting itself in a position to come, as quickly as possible, with proactive interventions to address competitive problems caused by platform giants just as soon as it determines it can legally do that.

The Bundeskartellamt also has a pioneering case against Facebook’s ‘superprofiling’ on its desk — which links privacy abuse to competition concerns and could drastically limit the tech giant’s ability to profile users. That investigation and case has been ongoing for years but was recently referred to Europe’s top court for an interpretation of key legal questions.

 

Indonesian healthcare startup Prixa has raised $3 million led by MDI Ventures and the Trans-Pacific Technology Fund (TPTF), with participation from returning investors including Siloam Hospitals Group.

This brings Prixa’s total raised to $4.5 million since it launched in 2019. Co-founder and chief executive officer James Roring M.D., told TechCrunch in an email that the new funding will enable Prixa to scale its platform and customer base. Prixa uses a B2B model, partnering with healthcare payers like insurance providers and corporations. Through its B2B customers, it currently serves about 10 million patients.

Prixa currently works with four major insurers and has six additional insurers in its short-term pipeline. It also works with Indonesia’s largest third-party administrators, Roring said, allowing it to reach more policyholders.

Prixa’s platform includes a digital health assistant to answer patients’ questions, telemedicine consultations, pharmacy deliveries and on-demand lab diagnostics. Usage increased during the COVID-19 pandemic as more patients sought online consultations for primary care.

Other telehealth startups in Indonesia include Halodoc and Alodokter (which is also backed by MDI). Both connect patients directly with healthcare and insurance providers. Roring said Prixa differentiates by focusing on greater cost control for healthcare payers and positioning itself as a digital primary care platform.

“By symptomatically managing patients outside of tertiary care facilities and caring for chronic non-communicable diseases online, Prixa is able to effectively reduce the amount of outpatient claims and downstream inpatient cost incurred by healthcare payers,” Roring said. “Additionally, the combination of a growing and robust medical database, as well as proven clinical guidelines, contribute to cost efficiency and service optimization through the standardization of treatment by our healthcare providers.”

In press statement about the funding, Aditia Henri Narendra, MDI Ventures’ general manager of legal and corporate communication, said, “MDI co-led this financing because Prixa has demonstrated its ability to support insurance companies and hospitals in making medical services more accessible and affordable through its AI telemedicine platform.”

An email has been going around the internet as a part of a release of documents related to Apple’s App Store based suit brought by Epic Games. I love this email for a lot of reasons, not the least of which is that you can extrapolate from it the very reasons Apple has remained such a vital force in the industry for the past decade. 

The gist of it is that SVP of Software Engineering, Bertrand Serlet, sent an email in October of 2007, just three months after the iPhone was launched. In the email, Serlet outlines essentially every core feature of Apple’s App Store — a business that brought in an estimated $64B in 2020. And that, more importantly, allowed the launch of countless titanic internet startups and businesses built on and taking advantage of native apps on iPhone.

Forty five minutes after the email, Steve Jobs replies to Serlet and iPhone lead Scott Forstall, from his iPhone, “Sure, as long as we can roll it all out at Macworld on Jan 15, 2008.”

Apple University should have a course dedicated to this email. 

Here it is, shared by an account I enjoy, Internal Tech Emails, on Twitter. If you run the account let me know, happy to credit you further here if you wish:

First, we have Serlet’s outline. It’s seven sentences that outline the key tenets of the App Store. User protection, network protection, an owned developer platform and a sustainable API approach. There is a direct ask for resources — whoever we need in software engineering — to get it shipped ASAP. 

It also has a clear ask at the bottom, ‘do you agree with these goals?’

Enough detail is included in the parentheticals to allow an informed reader to infer scope and work hours. And at no point during this email does Serlet include an ounce of justification for these choices. These are the obvious and necessary framework, in his mind, for accomplishing the rollout of an SDK for iPhone developers. 

There is no extensive rationale provided for each item, something that is often unnecessary in an informed context and can often act as psychic baggage that telegraphs one of two things:

1. You don’t believe the leader you’re outlining the project to knows what the hell they’re talking about.
2. You don’t believe it and you’re still trying to convince yourself. 

Neither one of those is the wisest way to provide an initial scope of work. There is plenty of time down the line to flesh out rationale to those who have less command of the larger context. 

If you’re a historian of iPhone software development, you’ll know that developer Nullriver had released Installer, a third-party installer that allowed apps to be natively loaded onto iPhone, in the summer of 2007. Early September, I believe. It was followed in 2008 by the eventually far more popular Cydia. And there were developers that August and September already experimenting with this completely unofficial way of getting apps on the store, like the venerable Twitterific by Craig Hockenberry and Lights Off by Lucas Newman and Adam Betts.

Though there has been plenty of established documentation of Steve being reluctant about allowing third-party apps on iPhone, this email establishes an official timeline for when the decision was not only made but essentially fully formed. And it’s much earlier than the apocryphal discussion about when the call was made. This is just weeks after the first hacky third-party attempts had made their way to iPhone and just under two months since the first iPhone jailbreak toolchain appeared. 

There is no need or desire shown here for Steve to ‘make sure’ that his touch is felt on this framework. All too often I see leaders that are obsessed with making sure that they give feedback and input at every turn. Why did you hire those people in the first place? Was it for their skill and acumen? Their attention to detail? Their obsessive desire to get things right?

Then let them do their job. 

Serlet’s email is well written and has the exact right scope, yes. But the response is just as important. A demand of what is likely too short a timeline (the App Store was eventually announced in March of 2008 and shipped in July of that year) sets the bar high — matching the urgency of the request for all teams to work together on this project. This is not a side alley, it’s the foundation of a main thoroughfare. It must get built before anything goes on top. 

This efficacy is at the core of what makes Apple good when it is good. It’s not always good, but nothing ever is 100% of the time and the hit record is incredibly strong across a decade’s worth of shipped software and hardware. Crisp, lean communication that does not coddle or equivocate, coupled with a leader that is confident in their own ability and the ability of those that they hired means that there is no need to bog down the process in order to establish a record of involvement. 

One cannot exist without the other. A clear, well argued RFP or project outline that is sent up to insecure or ineffective management just becomes fodder for territorial games or endless rounds of requests for clarification. And no matter how effective leadership is and how talented their employees, if they do not establish an environment in which clarity of thought is welcomed and rewarded then they will never get the kind of bold, declarative product development that they wish. 

All in all, this exchange is a wildly important bit of ephemera that underpins the entire app ecosystem era and an explosive growth phase for Internet technology. And it’s also an encapsulation of the kind of environment that has made Apple an effective and brutally efficient company for so many years. 

Can it be learned from and emulated? Probably, but only if all involved are willing to create the environment necessary to foster the necessary elements above. Nine times out of ten you get moribund management, an environment that discourages blunt position taking and a muddy route to the exit. The tenth time, though, you get magic.

And, hey, maybe we can take this opportunity to make that next meeting an email?

On-demand grocery delivery, which really came into its own with the emergence of the Covid-19 pandemic, continues to command huge attention from investors. The jury is still out on how people will use those services in the longer term, but in the meantime, the most ambitious of the startups in the field are raising big.

In the latest development, Flink — a Berlin-based on-demand “instant” grocery delivery service built around self-operated dark stores and a smaller assortment (2,400 items) of items that it says it will deliver in 10 minutes or less — has raised $240 million to expand its business into more cities, and more countries, on the heels of strong demand.

Flink — which means “quick” in German — is currently active in 24 cities across Germany, France and the Netherlands. It hasn’t disclosed how many active customers it has, but it targets younger consumers, those with small fridges, those who have forgotten items in their bigger shops, and people who simply don’t want to or can’t shop in the old-style of once every one or two weeks.

“We are on a mission to give people back some of their valuable time during their hectic days and impress them with our service every time they order,” said Flink CEO Oliver Merkel — who co-founded the company with Julian Dames and Christoph Cordes — in a statement. “We want to establish Flink as the top destination for their day-to-day goods at great prices and with instant delivery by our amazing riders. The order growth we have seen over the past weeks has been explosive and we attribute that to the excellent service we are providing to our consumers.”

The size of this all-equity Series A is extraordinary considering that company only launched in December last year. The company is not disclosing its valuation but one person close to the company said it’s “not a unicorn yet.” (Not worth $1 billion on paper, that is.)

The round is being co-led by Prosus, BOND, and Mubadala Capital, and it comes with a very interesting deal attached. REWE — a German supermarket giant — has inked a strategic partnership with the company that will make Flink its preferred partner for smaller shopping grabs, which looks like it will complement the work that REWE is doing to build out its own grocery delivery businesses for bigger baskets. It’s not clear if REWE is actually investing.

This latest investment comes on the heels of Flink announcing, back in March when it was only three months old, a $52 million round from Target Global and earlier backers Northzone, Cherry Ventures and TriplePoint Capital, along with Cristina Stenbeck from Kinnevik, who invested in a personal capacity.

The opportunity for a new startup to get into the market for food — and in this case specifically grocery — delivery, is an interesting one at the moment. On one hand, we’ve been through a year where many cities across Europe have been under shelter in place orders, pushing many more people to turn on online food delivery to get essential things to their doors.

That is to say, demand — at least under current circumstances — has been more than proven out, with many of the biggest providers completely buckling under pressure with crashing sites, very few delivery slots available and many items out of stock on a too-regular basis.

On the other, it’s led to a huge profusion of companies swooping in to fill that gap.

There are other new players like Gorillas, another outfit out of Berlin, which has also been raising big money and has boasted its own $1 billion+ valuation (for what it’s worth: remember, this is all just on paper). Alongside those are also a rush of more mature startups like Glovo (which raised $528 million earlier this year), Kolonial ($265 million earlier this year), Everli and Rohlik (respectively, $100 million and $230 million rounds this spring),  as well as much bigger players like Ocado and of course the brick-and-mortar grocers who are investing big in their own operations.

And just earlier this morning, Getir out of Turkey, another fast-grocery startup that has been investing a lot in growth (its delivery bikes are visible to me every time I go outside at the moment here in London) announced a $550 million round at a $7.5 billion valuation — a piece of news that likely was one part of the calculus for Flink also announcing today.

And there are so many more I’m not mentioning here.

The big question will be whether the market can sustain all of this, and if not, what that will mean for all of these, and all of the money invested in the space. It’s not unlike some of the scramble that took place in restaurant delivery, where a big profusion of regional giants first started out and then started land grabs to pick up others to get better economies of scale, a process that eventually took the most well-capitalized of them global. All of that is still playing out, and in fact some of the biggest of the hot-food delivery companies, such as Deliveroo out of the UK, are also moving into grocery to better diversify.

In that regard, it’s very interesting to see Prosus in this round. The company — the tech giant that was divided out from the rest of Naspers some time ago to better focus investment and attention on the space (it holds a huge stake in Tencent, among other things) — really got burned last year when its long, hostile attempt to acquire Just Eat to combine it with its existing holdings in food delivery was left bobbing in the water after Just Eat instead eloped with Takeaway.

Since then, it’s been very proactive in using capital to plot out its own course. That’s included stakes in Swiggy in India, investing in that Kolonial round, and also today’s news backing Flink.

“The opportunity that exists for online grocery delivery is vast, with the grocery market in Germany alone expected to reach more than €300 billion in the coming years,” said Larry Illg, CEO of Food Delivery at Prosus, in a statement. “The past year has seen many new players entering the nascent market, vying to fulfill the increasing consumer demands. Flink comes to the market offering ultra-fast delivery of items, mostly under 10 minutes, getting consumers what they need almost immediately. Flink’s innovative tech-enabled logistics service combined with the expertise of the team, the quality of the partnerships they have quickly established and the pace of execution within Germany, has been nothing short of impressive.”

In its latest ambitious digital policy announcement, the European Union has proposed creating a framework for a “trusted and secure European e-ID” (aka digital identity) — which it said today it wants to be available to all citizens, residents and businesses to make it easer to use a national digital identity to prove who they are in order to access public sector or commercial services regardless of where they are in the bloc.

The EU does already have a regulation on electronic authentication systems (eIDAS), which entered into force in 2014, but the Commission’s intention with the e-ID proposal is to expand on that by addressing some of its limitations and inadequacies (such as poor uptake and a lack of mobile support).

It also wants the e-ID framework to incorporate digital wallets — meaning the user will be able to choose to download a wallet app to a mobile device where they can store and selectively share electronic documents which might be needed for a specific identity verification transaction, such as when opening a bank account or applying for a loan. Other functions (like e-signing) is also envisaged being supported by these e-ID digital wallets.

Other examples the Commission gives where it sees a harmonized e-ID coming in handy include renting a car or checking into a hotel. EU lawmakers also suggest full interoperability for authentication of national digital IDs could be helpful for citizens needing to submit a local tax declaration or enrolling in a regional university.

Some Member States do already offer national electronic IDs but there’s a problem with interoperability across borders, per the Commission, which noted today that just 14% of key public service providers across all Member States allow cross-border authentication with an e-Identity system, though it also said cross-border authentications are rising.

A universally accepted ‘e-ID’ could — in theory — help grease digital activity throughout the EU’s single market by making it easier for Europeans to verify their identity and access commercial or publicly provided services when travelling or living outside their home market.

EU lawmakers also seem to believe there’s an opportunity to ‘own’ a strategic piece of the digital puzzle here, if they can create a unifying framework for all European national digital IDs — offering consumers not just a more convenient alternative to carrying around a physical version of their national ID (at least in some situations), and/or other documents they might need to show when applying to access specific services, but what commissioners billed today as a “European choice” — i.e. vs commercial digital ID systems which may not offer the same high-level pledge of a “trusted and secure” ID system that lets the user entirely control who gets to sees which bits of their data.

A number of tech giants do of course already offer users the ability to sign in to third party digital services using the same credentials to access their own service. But in most cases doing so means the user is opening a fresh conduit for their personal data to flow back to the data-mining platform giant that controls the credential, letting Facebook (etc) further flesh out what it knows about that user’s Internet activity.

“The new European Digital Identity Wallets will enable all Europeans to access services online without having to use private identification methods or unnecessarily sharing personal data. With this solution they will have full control of the data they share,” is the Commission alternative vision for the proposed e-ID framework.

It also suggests the system could create substantial upside for European businesses — by supporting them in offering “a wide range of new services” atop the associated pledge of a “secure and trusted identification service”. And driving public trust in digital services is a key plank of how the Commission approaches digital policymaking — arguing that it’s a essential lever to grow uptake of online services.

However to say this e-ID scheme is ‘ambitious’ is a polite word for how viable it looks.

Aside from the tricky issue of adoption (i.e. actually getting Europeans to A) know about e-ID, and B) actually use it, by also C) getting enough platforms to support it, as well as D) getting providers on board to create the necessary wallets for envisaged functionality to pan out and be as robustly secure as promised), they’ll also — presumably — need to E) convince and/or compel web browsers to integrate e-ID so it can be accessed in a streamlined way.

The alternative (not being baked into browsers’ UIs) would surely make the other adoption steps trickier.

The Commission’s press release is fairly thin on such detail, though — saying only that: “Very large platforms will be required to accept the use of European Digital Identity wallets upon request of the user.”

Nonetheless, a whole chunk of the proposal is given over to discussion of “Qualified certificates for website authentication” — a trusted services provision, also expanding on the approach taken in eIDAS, which the Commission is keen for e-ID to incorporate in order to further boost user trust by offering a certified guarantee of who’s behind a website (although the proposal says it will be voluntary for websites to get certified).

The upshot of this component of the proposal is that web browsers would need to support and display these certificates, in order for the envisaged trust to flow — which sums to a whole lot of highly nuanced web infrastructure work needed to be done by third parties to interoperate with this EU requirement. (Work that browser makers already seem to have expressed serious misgivings about.)

Another big question-mark thrown up by the Commission’s e-ID plan is how exactly the envisaged certified digital identity wallets would store — and most importantly safeguard — user data. That very much remains to be determined, at this nascent stage.

There’s discussion in the regulation’s recitals, for example, of Member States being encouraged to “set-up jointly sandboxes to test innovative solutions in a controlled and secure environment in particular to improve the functionality, protection of personal data, security and interoperability of the solutions and to inform future updates of technical references and legal requirements”.

And it seems that a range of approaches are being entertained, with recital 11 discussing using biometric authentication for accessing digital wallets (while also noting potential rights risks as well as the need to ensure adequate security):

European Digital Identity Wallets should ensure the highest level of security for the personal data used for authentication irrespective of whether such data is stored locally or on cloud-based solutions, taking into account the different levels of risk. Using biometrics to authenticate is one of the identifications methods providing a high level of confidence, in particular when used in combination with other elements of authentication. Since biometrics represents a unique characteristic of a person, the use of biometrics requires organisational and security measures, commensurate to the risk that such processing may entail to the rights and freedoms of natural persons and in accordance with Regulation 2016/679.

In short, it’s clear that underlying the Commission’s big, huge idea of a unified (and unifying) European e-ID is a complex mass of requirements needed to deliver on the vision of a secure and trusted European digital ID that doesn’t just languish ignored and unused by most web users — some highly technical requirements, others (such as achieving the sought for widespread adoption) no less challenging.

The impediments to success here certainly look daunting.

Nonetheless, lawmakers are ploughing ahead, arguing that the pandemic’s acceleration of digital service adoption has shown the pressing need to address eIDAS’ shortcomings — and deliver on the goal of “effective and user-friendly digital services across the EU”.

Alongside today’s regulatory proposal they’ve put out a Recommendation, inviting Member States to “establish a common toolbox by September 2022 and to start the necessary preparatory work immediately” — with a goal of publishing the agreed toolbox in October 2022 and starting pilot projects (based on the agreed technical framework) sometime thereafter.

“This toolbox should include the technical architecture, standards and guidelines for best practices,” the Commission adds, eliding the large cans of worms being firmly cracked open.

Still, its penciled in timeframe for mass adoption — of around a decade — does a better job of illustrating the scale of the challenge, with the Commission writing that it wants 80% of citizens to be using an e-ID solution by 2030.

The even longer game the bloc is playing is to try to achieve digital sovereignty so it’s not beholden to foreign-owned tech giants. And an ‘own brand’, autonomously operated European digital identity does certainly align with that strategic goal.

Toyota AI Ventures, Toyota’s standalone venture capital fund, has dropped the “AI” and is reborn as, simply, Toyota Ventures. The fund is commemorating its new identity by investing an additional $300 million in emerging technologies and carbon neutrality via two early-stage funds: the Toyota Ventures Frontier Fund and the Toyota Ventures Climate Fund. 

The introduction of these two new funds, each worth $150 million, brings Toyota Ventures’ total assets under management to over $500 million. With the new capital infusion into the Frontier Fund comes an expansion of Toyota Ventures’ core thesis, which previously focused on AI, autonomy, mobility, robotics and the cloud, and now is adding smart cities, digital health, fintech and energy. So while Toyota Ventures’ investment approach isn’t changing, it’s broadening the scope of startups it will consider investing in. 

“AI is kind of shrinking as a proportion of everything,” Jim Adler, founding managing director of Toyota Ventures, told TechCrunch. “The first mission of the Frontier Fund has always been to discover what’s next for Toyota. Toyota pivoted to cars in the 1930s, and Toyota will grow to other businesses in the future. Startups are experiments in the marketplace, and this is a way for us to understand and get comfortable with where innovations are coming from.” 

Toyota as a global company has more than 370,000 employees that cover a range of business units in which the company at large stands to benefit from investing, such as financial technology. The Frontier Fund is a step outside of mobility. It not only seeks to bring emerging tech to market, but it also wants to bring new innovations onboard, whether as a customer or an acquisition, according to Adler. 

“I think the vision of the company really is that machines are here to stay, they amplify the human experience, and Toyota understands how machines amplify humans really well for the benefit of society, which sounds incredibly corny, but the company really believes that,” said Adler.

By that same token, the new Climate Fund seeks to invest in startups that can help Toyota accelerate its goal of reaching carbon neutrality by 2050. The company has been investing in hydrogen for years, including a recent partnership with Japanese fuel company ENEOS, but it’s open to whatever technology will help achieve carbon neutrality, according to Adler.

“We think renewable energies will play a role,” said Adler. “Hydrogen production, storage distribution and utilization will play a role. We think carbon capture and storage will play a role. We’re not going to get dogmatic about hydrogen because we’ve been at it for decades and maybe things will change. Hydrogen hasn’t been crowdsourced across the startup community because there just wasn’t a market for it, but I think the market may be emerging.”

The fund is accepting online pitches on its website from entrepreneurs seeking early-stage funding. On Thursday, Toyota Ventures also announced it would be expanding its team and working with a new Advisor Network as a resource for founders looking for guidance on anything from product development to diversity and recruitment. 

“Toyota Ventures has been an invaluable partner for Boxbot since they invested in our seed round in 2018,” said Austin Oehlerking, co-founder and CEO of Boxbot, in a statement. “They have been instrumental in helping us to navigate complicated, existential challenges on our journey from concept to product/market fit. Jim and the team really understand how corporate venture capital should function in order to successfully partner with startups.” 

Adler says he and his team come from an entrepreneurial background, so they understand what it’s like on the other side of the table. Toyota Ventures’ focuses on early-stage startups because that’s where it believes some of the most interesting innovations come from. 

“I’m a big believer that early-stage venture capital is a telescope into the future,” said Adler. “I think we can actually find those incredibly valuable innovations that make this all worthwhile.”

A change to TikTok’s U.S. Privacy Policy on Wednesday introduced a new section that says the social video app “may collect biometric identifiers and biometric information” from its users’ content. This includes things like “faceprints and voiceprints,” the policy explained. Reached for comment, TikTok could not confirm what product developments necessitated the addition of biometric data to its list of disclosures about the information it automatically collects from users, but said it would ask for consent in the case such data collection practices began.

The biometric data collection details were introduced in the newly added section, “Image and Audio Information,” found under the heading of “Information we collect automatically” in the policy.

This is the part of TikTok’s Privacy Policy that lists the types of data the app gathers from users, which was already fairly extensive.

The first part of the new section explains that TikTok may collect information about the images and audio that are in users’ content, “such as identifying the objects and scenery that appear, the existence and location within an image of face and body features and attributes, the nature of the audio, and the text of the words spoken in your User Content.”

While that may sound creepy, other social networks do object recognition on images you upload to power accessibility features (like describing what’s in an Instagram photo, for example), as well as for ad targeting purposes. Identifying where a person and the scenery is can help with AR effects, while converting spoken words to text helps with features like TikTok’s automatic captions.

The policy also notes this part of the data collection is for enabling “special video effects, for content moderation, for demographic classification, for content and ad recommendations, and for other non-personally-identifying operations,” it says.

The more concerning part of the new section references a plan to collect biometric data.

It states:

We may collect biometric identifiers and biometric information as defined under US laws, such as faceprints and voiceprints, from your User Content. Where required by law, we will seek any required permissions from you prior to any such collection.

The statement itself is vague, as it doesn’t specify whether it’s considering federal law, states laws, or both. It also doesn’t explain, as the other part did, why TikTok needs this data. It doesn’t define the terms faceprints or voicepints. Nor does it explain how it would go about seeking the “required permissions” from users, or if it would look to either state or federal laws to guide that process of gaining consent.

That’s important because as it stands today, only a handful of U.S. states have biometric privacy laws, including Illinois, Washington, California, Texas, and New York. If TikTok only requested consent, “where required by law,” it could mean users in other states would not have to be informed about the data collection.

Reached for comment, a TikTok spokesperson could not offer more details on the company’s plans for biometric data collection or how it may tie in to either current or future products.

“As part of our ongoing commitment to transparency, we recently updated our Privacy Policy to provide more clarity on the information we may collect,” the spokesperson said.

The company also pointed us to an article about its approach to data security, TikTok’s latest Transparency Report, and the recently launched privacy and security hub, which is aimed at helping people better understand their privacy choices on the app.

The biometric disclosure comes at a time when TikTok has been working to regain the trust of some U.S. users.

Under the Trump administration, the federal government attempted to ban TikTok from operating in the U.S. entirely, calling the app a national security threat because of its ownership by a Chinese company. TikTok fought back against the ban and went on record to state it only stores TikTok U.S. user data in its U.S. data centers and in Singapore.

It said it has never shared TikTok user data with the Chinese government nor censored content, despite being owned by Beijing-based ByteDance. And it said it would never do so, if asked.

Though the TikTok ban was initially stopped in the courts, the federal government appealed the rulings. But when President Biden took office, his administration put the appeal process on hold as it reviewed the actions taken by his predecessor. And although Biden has, as of today, signed an executive order to restrict U.S. investment in Chinese firms linked to surveillance, his administration’s position on TikTok remains unclear.

It is worth noting, however, that the new disclosure about biometric data collection follows a $92 million settlement in a class action lawsuit against TikTok, originally filed in May 2020, over the social media app’s violation of Illinois’ Biometric Information Privacy Act. The consolidated suit included more than 20 separate class filed against TikTok over the platform’s collection and sharing of the personal and biometric information without user consent. Specifically, this involved the use of facial filter technology for special effects.

In that context, TikTok’s legal team may have wanted to quickly cover themselves from future lawsuits by adding a clause that permits the app to collect personal biometric data.

The disclosure, we should also point out, has only been been added to the U.S. Privacy Policy, as other markets like the E.U. have stricter data protection and privacy laws.

The new section was part of a broader update to TikTok’s Privacy Policy, which included other changes both large and small, ranging from corrections of earlier typos to revamped or even entirely new sections. Most of these tweaks and changes could be easily explained, though — like new sections that clearly referenced TikTok’s e-commerce ambitions or adjustments aimed at addressing the implications of Apple’s App Tracking Transparency on targeted advertising.

In the grand scheme of things, TikTok still has plenty of data on its users, their content, and their devices, even without biometric data.

For example, TikTok policy already stated it automatically collects information about users’ devices, including location data based on your SIM card and IP addresses and GPS, your use of TikTok itself and all the content you create or upload, the data you send in messages on its app, metadata from the content you upload, cookies, the app and file names on your device, battery state, and even your keystroke patterns and rhythms, among other things.

This is in addition to the “Information you choose to provide,” which comes from when you register, contact TikTok or upload content. In that case, TikTok collects your registration info (username, age, language, etc.), profile info (name, photo, social media accounts), all your user-generated content on the platform, your phone and social network contacts, payment information, plus the text, images and video found in the device’s clipboard. (TikTok, as you may recall, got busted by Apple’s iOS 14 feature that alerted users to the fact that TikTok and other apps were accessing iOS clipboard content. Now, the policy says TikTok “may collect” clipboard data “with your permission.”)

The content of the Privacy Policy itself wasn’t of immediate concern to some TikTok users. Instead, it was the buggy rollout.

Some users reported seeing a pop-up message alerting them to the Privacy Policy update, but the page was not available when they tried to read it. Others complained of seeing the pop-up repeatedly. This issue doesn’t appear to be universal. In tests, we did not have an issue with the pop-up ourselves.

Additional reporting by Zack Whittaker

The Biden administration has replaced and expanded Trump-era restrictions on investing in certain Chinese firms deemed supportive of that country’s surveillance and military apparatus. Major tech, space, and telecom companies are listed in the initial 59 covered by the Executive Order, with more to come by order of the Treasury.

“I find that the use of Chinese surveillance technology outside the PRC and the development or use of Chinese surveillance technology to facilitate repression or serious human rights abuse, constitute unusual and extraordinary threats,” writes President Biden in the introduction to the order.

The E.O. has its roots in the Trump administration’s long-running and evolving blacklist of Chinese companies, whether for government procurement, private investment by U.S. firms, or other purposes. Major tech companies like ZTE and Huawei were put on the list straight away in 2019, but others were steadily added over time.

The Biden order refines these orders, revising certain portions and expanding others, particularly in the definition of what constitutes dangerous behavior or collaboration with Chinese authorities. Notably it stretches this to cover companies involved in domestic surveillance of Uygur Muslims in China and political dissidents in Hong Kong and elsewhere.

The new list of companies includes many of those listed over the last two years, and adds plenty more. Seemingly any major company that deals with tech, communications or aerospace is at risk of being entered on the list, from China Mobile and China Unicom to China Aerospace, Hikvision, and SMIC. Direct investment in the companies is disallowed, as is investing in an intermediary such as an index fund that includes one of the prohibited companies.

The Treasury — rather than the Defense Department, as it was previously — is given the responsibility of maintaining and updating the list, either adding to or subtracting from it.

“Tackling these challenges head-on is consistent with the Biden Administration’s commitment to protecting core U.S. national security interests and democratic values, and the Administration will continue to update the list of PRC entities as appropriate,” read a fact sheet accompanying the order.

Clearly the White House aims to continue and refine the trade war with China started by Trump. Whether U.S. pressure will be enough to influence Chinese policy or if the international community’s support will be necessary may soon be clear as the President goes to visit allies in search of support on this and other measures.

Aurora, the autonomous vehicle company that acquired Uber ATG last year, has assembled a team of outside experts, shared new details about its operations in a self-assessment safety report and launched a website as part of a broader effort to win over consumers wary of the technology that they may someday share the road with, or even use.

Aurora said Thursday it has tapped experts in aviation safety, insurance, medicine and automotive safety — all people from outside of the niche AV industry — to provide an outside perspective on the company’s overall approach to safety, to look for gaps in its system and advise on the best ways to share its progress and record with regulators and the public. The advisory group is designed to augment Aurora’s existing safety efforts, which includes on-road testing and development.

“I think for a while we’ve almost done the ‘Field of Dreams’ analysis where it’s like, ‘well if we build it they will come, just look at iPhones,'” Nat Beuse, Aurora’s head of safety said in a recent interview with TechCrunch. “We are always comparing it to these other consumer products, and I’m not so sure that is actually how we win over the hearts and minds of consumers in every single community in the United States.”

Beuse, who previously led the safety team at Uber ATG and once oversaw automated-vehicle developments at the U.S. Department of Transportation, said the goal is for driverless vehicles — whether that’s robotaxis shuttling people or trucks hauling freight — to be adopted broadly. That can’t happen, he said, without being able to measure and show the public that the technology is safe. He noted that public trust is one of the two biggest threats he sees to the AV industry.

“If all we worry about is a small number of people who get exposed to [AVs] we will never see the benefits of this technology and the broad scale, sweeping changes and the impact that it can have on our lives in a beneficial way,” he said. “We have to do a lot more there [gaining public trust]. Beuse added that gaining public trust should be done in concert with the government.

“I think for too long it’s been, ‘You, industry, you solve it. You’re building this stuff,'” he said. “And I really think it’s a partnership. Of course, we’re building the tech, we have a huge responsibility, but also the government has a huge, huge role to play and helping us kind of get the public on board.”

The members of the safety advisory board include Intelligent Transportation Society of America President and CEO Shailen Bhatt, Dave Carbaugh, the former chief pilot for flight-operations safety at Boeing and Victoria Chibuogu Nneji, the lead engineer and innovation strategist at Edge Case Research. Other members include Biologue President Jeff Runge, who is also a former administrator of the National Highway Traffic Safety Administration, Adrian Lund, managing member of HITCH42, LLC and former president of the Insurance Institute for Highway Safety and GHS Aviation Group CEO George Snyder.

The committee, which has already been meeting, is comprised of people who “don’t live and breathe the tech,” Beuse said.

Most importantly, for Aurora and the rest of the industry, is addressing the looming question of ‘how safe is safe enough?’ when it comes to driverless vehicles. One metric that has been adopted, and increasingly criticized, is comparing vehicle miles traveled and vehicle miles per “disengagement,” an industry jargon term that means a human safety operator has taken over from the computer driving the vehicle.

“We’ve been pretty adamant that that’s not a real metric because you can drive around in a parking lot and generate some interactions and that’s a whole lot different than if you’re driving in a city — and oh by the way, that’s a whole lot different if you’re driving on the highway,” Beuse explained.

Aurora is part of the Automated Vehicle Safety Consortium (AVSC), which includes Daimler, Ford, GM, Honda, Lyft, Motional, SAE and Toyota, that is working to come up with better safety metrics. The new Aurora safety advisory board isn’t working directly on the AVSC project, however it is providing general guidance that could help in this effort.

While there is still more work to be done to validate these new metrics, the group does have a handful that it thinks are pretty promising, Beuse said.